Articles by Irene Serrano

How Sampler Reduces Time to Insights by 98% [Case Study]

Irene Serrano — Fri, 04 Dec 2020 14:18:06 +0000

With the right user-friendly analysis tools, companies can access the meat of consumer reviews and understand the sentiment behind why customers feel the way they do about a product. Of course, before anyone can write a review or even give a product a star rating, they have to try out the product first. Sampler is streamlining that process with its user profiles, personalized samples and doorstep delivery.

1. Why Sampler Needed Sentiment Analysis for Product Review Data

Sampler works on the simple logic that if consumers have the right products in their hands, they’re more likely to use those products. They also found that 1 in 4 consumers who get a free sample leave a review of the product. Taking into account a decrease in the number of avenues for consumers to trial products or purchase samples due to the global pandemic, this kind of personalization of sampling combined with delivery to customers’ doors boosts both metrics.

When she collected client customers’ ratings and reviews, it was simple enough to analyze the ratings, but analyzing anecdotal text-based feedback and relaying it to clients wasn’t in the cards.

Sampler account manager Alisha Manion found that it was difficult for her clients to sift through thousands of reviews and determine any real trends. When she collected client customers’ ratings and reviews, it was simple enough to analyze the ratings, but analyzing anecdotal text-based feedback and relaying it to clients wasn’t in the cards.

Some Sampler programs gather as many as 50,000 reviews. That’s where being able to analyze reviews at scale has been extremely helpful for the growing company. Crafted to detect and analyze sentiment within large quantities of text-based qualitative data, Keatext’s AI tool gives businesses like Sampler added value to their clients.

“When there’s a rating and review component to the reports that I present to a client, I can pull out the top topics that have been mentioned and present that information to the client,” says Alisha. “I can report that in a graph so they can see very clearly what major themes are emerging and what people are really gravitating towards.”

2. How Keatext’s Sentiment Analysis Provided ROI for Sampler

For Sampler, the primary ROI with Keatext’s review analysis solution is in how the tool helps account managers like Alisha quickly pull data and find the insights that matter most to clients.

“Instead of 8 hours to parse the details of the data in 6,000 reviews, it takes me 10 minutes using Keatext”

“Even reading through a few hundred ratings and reviews to see what jumps out and what some of the themes are takes a lot of time that we don’t have, especially in a pandemic when our company has grown quite a bit,” Alisha explains. “Instead of 8 hours to parse the details of the data in 6,000 reviews, it takes me 10 minutes using Keatext.”

For example, if Keatext identifies “scent” as a major topic customers are mentioning, Alisha can see how often scent is mentioned as a positive or a negative sentiment—often the review also mentions the reason behind that sentiment. “Being able to use even just the top level of Keatext analysis to pull out those nuances has been useful for understanding customer sentiment,” she adds.

Using sentiment analysis on unstructured data, Keatext’s AI tool can show multiple classifications within customers’ reviews, distinguishing between subjective and objective, opinion and fact, negative and positive, comparative and direct, and so on. The insights gained through this deep level of analysis can help uncover the motivation behind customer engagement with a brand.

3. Which Sentiment Analysis Features Mattered to Sampler

For Sampler account managers and other Keatext users with specific client needs to meet, Keatext’s “topic” view lets them group relevant topics together to include in reports on product roll outs and other campaigns. Having that level of control over the platform as a user makes customer data even more relevant to how every user does their work.

“Adding insights from data analysis to our reports, pulling out key topics and comments, it gives clients more context around consumer reviews and really shows that feedback’s value to a brand”

“Adding insights from data analysis to our reports, pulling out key topics and comments, it gives clients more context around consumer reviews and really shows that feedback’s value to a brand,” says Alisha.

Keatext lets all clients build custom dashboards so they can see the data and insights that matter most to them and their needs. These dashboards feature a summary of all the comments analyzed; top positive and negative comments, topics and sentiments; as well as statistics of the data and related graphs—all based on the KPIs that a company is interested in.

As the number of product reviews increases online, data analysis of customer feedback can help brands not only understand what their customers think about their products but use data-derived insights to brainstorm how to better engage with those customers. A free product sample delivered right to a customer’s doorstep is one way to reach out—when it’s coupled with data analysis of reviews that gives clients more insight into customer opinion and behaviour, the path to customer engagement becomes much more clear.

The post How Sampler Reduces Time to Insights by 98% [Case Study] appeared first on Keatext.

Data that pops: Keatext’s new statistics module illustrates every insight

Irene Serrano — Wed, 25 Nov 2020 16:44:17 +0000

Keatext’s statistics module allows users to stay in the Keatext platform to create their own charts based on key data points, patterns in the data, and the insights they want to highlight.

The process begins with a dataset of any size, uploaded by a Keatext user and analyzed by the platform’s AI data analysis tool. Keatext’s statistics module works on the principle of pivot tables, allowing users to configure the layout of the data findings in several ways, such as seeing the selected fields as rows in a hierarchical column, or displaying the results as a pivot view where the selected fields are in separate rows and columns with totals at the end of each row.

Making visual sense of data hierarchies with NLP sentiment analysis

With data laid out on the screen, the user can then access the tool’s Fields popup, where they can select the data they want to show in a grid or chart by dragging and dropping hierarchies from the Fields list to the Rows, Columns, Values, or Report Filters boxes.

This means that Keatext users can focus on the data that interests them most for a given purpose, such as particular details about who the data is based on, including location and age, or, considering Keatext’s focus on analyzing people’s verbatim feedback, certain emotion-based negative and positive sentiments within customer or employee feedback data. Cells from the data source can also be formatted for number formatting and conditional formatting.

Keatext’s pivot charts let users visualize data in a new way by using filters to drill down through data to better choose what to highlight in a graph or grid. As part of a platform created to reveal insights within data, this visualization tool enables users segment their data by field so that any viewer can understand that data in relation to the other fields it exists within—thereby rendering insights with more clarity. For example, the statistics module can produce a colour-coded chart that lets viewers compare how many people bought an item in several different cities and how their satisfaction levels might differ.

The colours of the visuals can be changed to a company’s brand colours and every chart or grid is easy to export into multiple formats.

The post Data that pops: Keatext’s new statistics module illustrates every insight appeared first on Keatext.

Turning Employee Voice into Business Value with a Leader in the Canadian Telecom Market

Irene Serrano — Fri, 06 Nov 2020 18:17:05 +0000

The modern workplace has changed, creating a new strategic landscape for HR leaders who are now expected to guide their companies through constant change, retain top talent, and demonstrate the ROI of their work. In this new context, listening to employees is the single best way to build a strong, successful company.

One of Canada’s biggest telecom companies is no stranger to this monumental change. The team at Keatext analyzed thousands of public employee reviews from the company to demonstrate how Keatext turns employee feedback into clear insights that HR leaders can use to make a real difference.

The Challenge: High Turnover, Low Trust

Turnover and retention are the number one concerns for HR leaders today. Hiring budgets are tight, and with a quarter of the workforce likely looking for a new job, keeping and developing existing talent is essential to save on hiring costs. To do that, every company needs a culture where people come first. But that’s hard to build when trust is low and burnout is high.

For many companies, only half of their employees actually trust their leaders, and three out of four feel burned out. These are not just surface problems but deeper root causes that affect the bottom line. To fix them, leaders need to know what their employees are really thinking. HR teams already understand that an employee listening program is how to get there. It’s a straightforward way to gather, understand, and act on feedback to improve engagement, build trust, and keep their best people.

The challenge is turning all that feedback into actions and outcomes that demonstrate ROI. Executives want HR teams to automate and cut costs, but teams are already stretched thin. As much as 47% of HR leaders are uncertain about how to drive change. They need a partner like Keatext who understands their challenges and provides the right technology to overcome them.

The Approach: Deeper Data Analysis

To get a real, unfiltered look at the employee experience at this telecom company, the team at Keatext took 1,698 employee reviews from Indeed.ca over a period of seven years, a dataset that included 3,720 comments and 7,592 individual statements. This kind of unsolicited, unbiased feedback is an invaluable resource for understanding what really matters to employees.

Keatext’s AI takes all that raw feedback and turns it into clear insights that are immediately actionable. For HR teams, this capability delivers a critical advantage. It gets the team past basic sentiment analysis and identification of top themes into the “why” behind employee feedback that provides a solid, data-driven foundation to guide strategic decisions.

The Solution: What the Data Told Us

Keatext’s analysis gave a clear picture of what it’s like to work at this telecom company.

The good news? Most employees are happy. 65% gave the company a 4 or 5-star rating, and Keatext’s AI found 2,460 positive comments compared to 1,124 negative ones, plus a handful of suggestions and questions.

The bad news? Satisfied employees still have issues—and there’s a fine line between overall satisfaction and unresolved issues that eventually lead to turnover.

Ultimately, the real value is in the details. The top themes that employees mentioned were Service, Training, People, Job, and Environment. Keatext’s AI also identified a spike in negative feedback over a period of a few months. Upon looking into the comments from that time, the team uncovered employee dissatisfaction with the number of good managers and training for sales reps.

This problem directly hurt team morale and spilled over to customers, who got bad advice and were unhappy with the service. The data showed a clear link between the employee experience and the customer experience.

From Insight to Impact: Solving Real HR Problems

Keatext gives HR leaders the clarity to tackle their biggest challenges head-on, turning insights into concrete action with AI-driven recommendations.

Supporting Employee Wellbeing to Prevent Burnout

A continuous listening strategy is essential for proactively addressing issues like burnout. The data showed that employees found the work intense and the hours long, directly impacting their wellbeing.

This is where employee listening moves from passive data collection to active problem-solving. Keatext’s AI identifies the root causes of burnout and generates specific recommendations, such as proposing frameworks for flexible scheduling or providing data-backed arguments for manager training focused on work-life balance. This allows HR to move from problem identification to solution-building instantly.

Rebuilding Trust Through Better Management

Effective employee listening builds trust, and the data proved that management was a key area where trust was eroding. Employees consistently brought up issues with “middle management” and a “lack of quality managers.”

Keatext acts as a partner in this process, analyzing feedback on topics like “manager” and “communication” to pinpoint where support is needed. The platform then recommends targeted interventions, from leadership coaching modules to strategies for improving team feedback channels. This approach shows employees that their feedback is heard and acted upon, a cornerstone of rebuilding trust.

Retaining Talent by Focusing on Growth

A smart employee listening program looks beyond immediate frustrations to understand long-term career aspirations. The analysis revealed concerns about the “difficulty in career advancement,” a major reason people leave their jobs.

Keatext revealed where those roadblocks are by analyzing comments about “promotion” and “career growth.” The platform’s AI can then recommend tailored professional development strategies or suggest structural changes to create clearer career paths. This is a best practice for using employee feedback to foster a more engaged and committed workforce.

Connecting Employee Experience to Business Outcomes

Employee wellbeing and customer satisfaction are deeply connected. A holistic employee listening strategy recognizes this. The link between undertrained sales reps and unhappy customers is a powerful insight that proves this connection.

Keatext provides concrete, AI-driven recommendations to solve these internal issues, helping HR leaders demonstrate the clear business case for investing in employee wellbeing. It connects the dots between a supportive internal culture and measurable gains in customer satisfaction and profitability.

The Outcome: A Strategic Partner for Data-Driven HR

Keatext acts as a strategic partner to turn HR into a core function that drives the business forward.

AI-Powered Analysis: Reading through thousands of comments manually is no solution. Keatext uses smart AI to do the heavy lifting, automatically analyzing unstructured text to uncover sentiment, organize feedback into clear topics, and generate insights without the guesswork and human error of doing it by hand.
Clear Recommendations: Keatext helps leaders decide what to do next. Our platform gives AI-driven recommendations for specific areas—like manager support or work-life balance—so you can focus your efforts where they’ll make the biggest difference. Recommendations are generated using LLMs that consolidate knowledge into smart, contextual recommendations for your business.
Fits Right In with Existing Tools: Keatext is a cloud-based platform that works with existing HR tech. It can easily pull in data from surveys, 1-on-1 notes, focus groups, and even public review sites. It gives HR leaders one place to see the complete picture.

With Keatext, HR leaders are able to build and run a continuous feedback loop. They can ask for honest, open-ended feedback without worrying about how to analyze it. They can build a rock-solid case for putting their people first. And they can show their employees that their voice is heard and their wellbeing matters. That’s how to build a company where people trust their leaders, stick around, and do great work.

The post Turning Employee Voice into Business Value with a Leader in the Canadian Telecom Market appeared first on Keatext.

Coveo’s AI-powered search platform hits every touchpoint

Irene Serrano — Tue, 13 Oct 2020 19:43:38 +0000

In the past several months, retailers have seen how COVID-19 has accelerated numerous digital trends already set in motion before the pandemic, including higher customer expectations when it comes to searching for products and services online. Whether a simple website search, e-commerce shopping query or chatbot question about shipping, customers are often met with off-target answers that leave them frustrated or on the way out the digital door.

Solving the problem of search disengagement, new artificial intelligence tools let companies deliver the most relevant answers to their customers — and transform the bigger customer data puzzle in the process.

Solving the problem of search disengagement, new artificial intelligence tools let companies deliver the most relevant answers to their customers — and transform the bigger customer data puzzle in the process.

Helping companies integrate AI-powered search into CX, customer service, e-commerce and employee experience, Montreal-based Coveo taps into a goldmine of data to generate relevant and engaging search results at every touchpoint. Hand-in-hand, customer data and AI tools that make that data actionable lead to a more personalized and relevant retail customer experience, from a customer’s initial product search to upselling and purchasing.

“It’s that notion of bringing product and content together in the right space, in the right place in the customer journey,” says Mark Floisand, Coveo’s SVP Product & Industry Marketing. “Customers don’t want to sit in a call centre hold queue and wait for an answer that might or might not be right.”

Coveo builds their machine-learning models based on data that is the most meaningful to companies and their customers. So for companies with all kinds of valuable information strewn throughout their systems — product specifications, fixes, cheat sheets, troubleshooting guides, customer feedback and more — Coveo creates a common index from that data, turns it loose through search and returns results that customers are looking for.

From AI search utility to personalized CX

As with customer service-related searches, the search or chat box of an e-commerce or retail website is a dynamic way for customers to interact in their own words with a company, and provide companies with important information about their needs. The difference is that Coveo’s AI-powered search, with access to all the information on every product of a company, is able to understand a customer’s query and give them the details they need on the products they’re looking for, as well as related products they’re likely to be interested in. That strategy leads more effectively from brand engagement to increased sales.

Coveo’s AI-powered search, with access to all the information on every product of a company, is able to understand a customer’s query and give them the details they need on the products they’re looking for, as well as related products they’re likely to be interested in.

Considering the volume of data on products and customers today, true personalization of CX can only be made possible through AI tools. A successful desktop search utility 15 years ago, Coveo saw an opportunity in multi-tenant cloud-based architecture: the company modernized its platform with AI, making it easily deployable and able to track analytics data at scale.

In his work with tech giants from Apple and Adobe to Sitecore and SAP, Floisand has seen firsthand how AI needs to be used as a strategic tool for meeting customer expectations and achieving business objectives, in CX and beyond. Coveo has seen its platform increase companies’ search capacities and help them orchestrate more complex and successful customer journeys — the AI company was recently named as a strong performer in The Forrester Wave: Journey Orchestration Platforms, Q2 2020. It’s a level of choice and flexibility in CX that has quickly become what customers expect.

“In a perfect world, a marketer would say here’s my ideal customer journey map, how do I steer people through it?” explains Floisand. “That perfect world doesn’t exist though. The only right map is the one for that individual at the moment they’re in. If you funnel all these individuals through one journey path, many of them will resist or abandon it. So why not use machine learning to give you the best chance of getting the most relevant content to every individual, wherever they are on their own journey?”

Putting data to work for retail customer experience

In retail commerce, where personalization has to happen in a few clicks after a customer lands on a website, companies need AI tools that quickly put together those clicks as data points. Coveo’s AI collects data around those clicks and around the products themselves, builds a multi-dimensional, personalized e-commerce space out of that data, then provides relevant and up-to-date search results and other recommendations. Customers will stay on the site and move closer to a sale.

In that way, AI platforms like Coveo are helping customers get what they need while changing the experience of the user interface itself, an essential yet often overlooked part of the customer journey. For a customer who wants to buy a one-off, higher-ticket item like a freezer, they’ll typically search out a local retailer rather than go to a marketplace like Amazon. Browsing that smaller retailer’s online catalogue, however, isn’t always a pleasant experience, digging down through categories on a sidebar and reading extraneous details that might slow down or halt their purchasing decision. If a customer cares most about a freezer’s dimensions, Coveo’s tools make sure that’s what they see first.

“Coveo is doing this on the basis of proven outcomes, not only what people searched for but what purchase paths were successful,” says Floisand. That virtual circle of positive experience from companies to customers and back again all comes down to integrating AI tools into a CX strategy that itself is integrated into a data-oriented business plan.

Connecting a company’s most meaningful data

While most organizations use several system tools in multiple departments to solve specific issues, typically these systems don’t talk to each other. AI solutions can tie systems together through pockets of user data. By connecting, indexing and unifying data with AI, companies can serve more meaningful content to their departments and to the customers they serve — and keep data more secure on top of that. In aggregate, all that data provides a more meaningful view of customers and their choices.

While most organizations use several system tools in multiple departments to solve specific issues, typically these systems don’t talk to each other.

“The more we can avoid silos and utilize the commonality of interaction data, the better,” says Floisand. “If you’ve got a lot of traffic on an e-store or a self-service portal, the tool has learned a lot already about customer flow from interaction data. So if a company wants to spin up a chatbot service, for example, it can take advantage of where existing volume traffic has shown proven success and bring that to bear in a new mode.”

Floisand illustrates the process with an example of a customer searching for a notoriously hard-to-purchase-online product: tents. “People want to know not just their dimensions but how easy they are to put up, so they do a lot of scouting on YouTube. If a company has tent construction videos generated by them, their vendor or their customers, a tool like Coveo can bring that video content right onto the tent category page of a company’s website and offer other relevant products alongside it, actively helping someone make a decision. It’s going to improve conversion rates, it’s going to enable customers to buy what they want through an experience they enjoy, and make it less likely that they’ll return products.”

To an AI platform, the video is simply more data to index and present in the shopping flow. From a CX perspective, that indexed video data hits all the business metrics — increased conversion, higher value due to upselling along the way, lower return rate — and leads to greater profitability.

Helping retailers be retailers, not tech companies

The future of how people buy online will depend not only on new technologies but on strategic brand experiences and engaging e-commerce. Retailers and brands facing the rise of massive online retail marketplaces have to stand out from their competition and create experiences that drive demand — personalizing retail customer experience is one major way to differentiate themselves.

The future of how people buy online will depend not only on new technologies but on strategic brand experiences and engaging e-commerce.

“In e-commerce, 50% of the market is concentrated in the hands of four big players, but what about everybody else?” asks Ciro Greco, Director of AI at Coveo. A neuroscientist who transitioned to tech, Greco emphasizes the importance of AI in natural language processing to boost retailers’ search capabilities into the realm of personalized CX.

“Many companies in the retail commerce sector aren’t super technical and will generate data in a very different way than the big players,” Greco continues. “On top of that, most retailers just aren’t AI companies. Coveo is saying, ‘be a retailer, don’t try to be a tech company’. Get the right tools and focus on your products and revenue path.”

For most brands and retailers, what matters most in terms of data is not necessarily quantity but quality, since their customer base comes to them with specific needs. To stay competitive and generate revenue, these companies don’t need to deal in nearly as much data as big marketplace players do — but they do need smart data solutions that integrate into their systems, stay current and make their data actionable.

“Coveo’s broader function is as a uniform layer stretched across a company’s customer touchpoints,” explains Greco. Since every touchpoint has its own idiosyncrasies, a unified indexed layer of data can help align the goals of different departments, leading to better understanding of customer behaviour at every touchpoint. “When a customer lands on your website, they’re in a place and going somewhere,” Greco reiterates. “There’s a lot of information we can take out of that alone, and that’s exactly where the AI comes in.”

At the heart of Coveo’s platform is a drive to make companies more competitive through AI. “When there’s something that your company does very well, there’s no need to be wiped out by the big players,” says Greco. For traditional retailers who are rebuilding and augmenting their websites and online stores to meet customer expectations, AI tools like Coveo’s are a quick and effective way to put data to work bridging the gap between retail customer experience and successful business outcomes.

– written with files from Keatext contributor Robyn Fadden

The post Coveo’s AI-powered search platform hits every touchpoint appeared first on Keatext.

Heyday’s chatbot: Improving CX with conversational AI

Irene Serrano — Wed, 23 Sep 2020 14:28:34 +0000

Everyone who’s shopped online recently has experienced the chatbot pop-up: much like an in-store salesperson, the chatbot opens with a friendly “What can I help you with?” Many customers ignore it and keep browsing, others answer the question – sometimes they even receive the right answer. The similarities between a really great salesperson and a great chatbot come down to knowledge and demeanour. Until recently, the salesperson blew the chatbot out of the water on both counts.

Today the chatbot is not only catching up but rubbing virtual shoulders with sales and service agents to drive customer interaction, sales and satisfaction. In a COVID world shifting away from in-store experiences, new conversational AI technologies such as chatbots and voice assistants can be anywhere at any time, programmed to engage in more natural conversation, and capable of quickly connecting customers to the right service or sales people before they even contemplate dropping out of the chat.

Montreal-based AI company Heyday puts conversational AI into action in the rapidly changing retail sphere, where the customer journey has taken more than a few twists and turns lately. Heyday co-founder and CEO Steve Desjarlais recently outlined how the COVID pandemic has disrupted shopping habits, yet it hasn’t changed what customers want in a retail experience, whether it’s online or in-store: they want to have their questions answered, get the products or services they actually want, and along the way be treated as an individual.

“Shopping online isn’t just about convenience anymore. It’s evolved into something much bigger,” writes Desjarlais. In this case, something bigger really can be better, with more personalization throughout the customer journey, from customized search results to chatbots with the right answers and the ability to know when to transfer a customer to a human agent. The new normal has seen some brands rocket their digital strategies, cramming 10 years of evolution into a few months. It’s serendipitous that a boom in AI technology is happening at the same time, allowing companies who integrate AI tools into their CX to rapidly adapt to trends and grow.

Changing channels and the conversation

As a conversational AI platform, Heyday lets people communicate with websites, apps and devices simply by writing or talking as they naturally would. The difference with Heyday is its focus on not only recognizing human language and automating common interactions with ease, but definitively understanding what customers want, wherever they are in the CX journey.

“Heyday’s AI layer is on the front end of a brand’s website, there to discern the customer’s intent,” says Brad Wing, VP Strategy and Partnerships at Heyday. “That’s a really important position because 50-70% of questions can be the same, whether they sell bolts or watches or energy drinks: where’s my order, what’s going on with my refund, is the store open.”

The chatbot can meet customers where they are or where they prefer to be, physically or digitally, and can help serve them what they want or need based on past experiences, comparative audiences or the mining of deep data.

“Looking at common customer chat data, that boils down to about 150 individual questions, each of them having thousands of variations in multiple languages. A Heyday chatbot is able to handle up to 80% of FAQs coming in for customer service, and the remaining 20%, we escalate to the right team,” says Wing. “The chatbot can meet customers where they are or where they prefer to be, physically or digitally, and can help serve them what they want or need based on past experiences, comparative audiences or the mining of deep data. In terms of CX, that means customers can decide how they want to talk to the brand – whether over messaging, email, SMS on their phone, it’s up to them.”

For brands who want to more personally engage with customers, Heyday’s ecommerce and CRM integrations with Shopify, Lightspeed, Magento and Salesforce make it an attractive solution for both engagement at scale and monetization of customer service. The added bonus of Heyday’s AI layer is data captured from chats that can be brought into a company’s data strategy for understanding customer behaviour, boosting sales and maintaining meaningful customer relationships.

Coming from heavyweights in ecommerce (Lightspeed) and the video game industry (Ubisoft), the founders of Montreal-based AI company Heyday decided to build a better chatbot: they combined the sticky experience of a best-selling game with the latest in personalized CX. Heyday’s early pivot from the banking sphere to retail tapped into forward-thinking technological and consumer trends, solidifying the company’s vision for elevated, customized CX at scale.

“On the technology side, we saw more APIs become available and companies like Google released AI libraries that could be built on by other companies. At the same time, we saw a rise in consumers moving their conversations to communications tools like Facebook Messenger, WhatsApp, WeChat and other messaging channels,” says Wing.

“Heyday is the bridge between the ecommerce technology and the customer’s voice,” he explains. The channel from Heyday’s perspective could be Facebook Messenger, WhatsApp, Apple Business Chat, Google’s Business Messages or any text conversation channel that connects a company with its customers, helping brands have a conversation with their customers in the way that their customers want to communicate, rather than directing them into a phone call or an email.

AI for the people: How to improve customer experience with chatbots

Available as a self-serve app from the Shopify marketplace, the “mini version” of Heyday takes just 60 seconds to set up. Suddenly an artisan can have an AI chatbot on their storefront. What was almost unimaginable a year ago – a generic consumer-level AI chat tool – is now being used by hundreds of small businesses across the globe.

Recently, Heyday has leapt headfirst into democratization of its AI platform. While many companies want to take advantage of the platform’s power for customization and CRM integration, small companies and solo entrepreneurs now have access to Heyday through Shopify. Available as a self-serve app from the Shopify marketplace, the “mini version” of Heyday takes just 60 seconds to set up. Suddenly an artisan can have an AI chatbot on their storefront. What was almost unimaginable a year ago – a generic consumer-level AI chat tool – is now being used by hundreds of small businesses across the globe.

Meanwhile, the decidedly bespoke side of Heyday is giving medium to large companies like sporting goods retailer Decathlon and Make Up For Ever more flexibility in CX. These brands can integrate the AI chat technology to engage with customers across the channels they use most, from Facebook Messenger to Google Maps messages to WhatsApp.

“Our conversational platform sits in the middle of all this, collecting the information from those channels,” says Christine Dupuis, VP Product and Growth at Heyday. “The AI recognizes patterns and intents across each channel. For instance, users from certain demographics prefer Facebook, while Google Maps and Search inquiries are more often about store location or hours. Whatever they are about, whatever customers need, wherever they are, a brand can be there and be ready to answer.” With its attention to accurate automation, the AI conversational platform can answer up to 80% of support questions and direct the remaining 20% to a human agent.

“Because we have AI, we can provide open text Q&A for customers, letting people write whatever they want in the chat box,” says Dupuis. “Through a customer’s own words, we can see what their issue is, what they’re looking for and follow their journey. Brands don’t have to watch customers go down a predetermined path and wonder why they don’t want to follow it and drop out. We know that the customer journey is not as linear as it once was, so leaving it open – saying ‘You be you’ to customers and assuring them that the chatbot will figure out what they need – gives brands the opportunity to capture truly useful data that can be used to create a better customer journey.”

From conversation to increased sales

The line between customer service and sales blurs when an AI chatbot enters the CX space. Usually the employees managing a brand’s website and chat are support personnel rather than an expert in products or services. Conversational AI can be both. “Heyday’s platform makes that expertise available through a chat. It’s not meant to feel like a ticketing system of support, it’s meant to feel like talking to a person in the store,” says Dupuis.

Today every connection, every touchpoint is so important that you can’t let your customer service person only handle customer service problems.

High-growth ecommerce companies are harnessing this advantage and seeing close to a 50-50 split between customer service questions and sales questions in their customer-focused chat platforms. “Today every connection, every touchpoint is so important that you can’t let your customer service person only handle customer service problems,” remarks Wing. “For retail especially, it’s becoming a hybrid touchpoint: the agent can talk about sending a shipping label with a customer but also delve into why they need to return an item, for instance. Find the answer to that, you can save a sale.”

With a brand’s catalogue integrated into the backend of the AI platform and a conversational design team crafting a chat experience that feels authentic and engaging, more customers than ever are also using the chatbot as a search tool to find the products they want – and more beyond that.

“Though the way people shop has changed, especially with COVID, we’re still focusing on creating a pleasant, natural, positive experience,” says Dupuis. “Some of our retail clients now offer appointments to customers who prefer to shop in-store, so Heyday is adding appointment scheduling to the chat. In response to increased demand in the luxury market and from other big-ticket brands, we’ve started to add video calls to the chat, letting sales people and stylists easily show items to their customers before buying. In the future, as brands continue their journey to hyper-personalized experiences, our platform will increase its ability to push personalized product recommendations based both on the users intent and increasingly their context and history.”

One of Heyday’s newest features is also one of AI technology’s strengths when it comes to deriving actionable insights from massive amounts of data: the chatbot can detect sentiment in customers’ written text. “Most retailers see the advantage of still having a conversation, of having empathy with a customer in that space,” says Wing. “An AI conversation platform gives brands another avenue to get that message across and say ‘we hear you, we’re going to make every effort to listen to you and give you options you need’.”

“In a website chat box, we’re reproducing the experience that customers would have if they went into a store and spoke with somebody there,” says Dupuis, clarifying that though chat can’t be a complete duplicate of the in-person experience, of course, Heyday’s chat solution gets customers the answers they actually need, then and there, wherever in the digital world customers may be.

– written with files from Keatext contributor Robyn Fadden

The post Heyday’s chatbot: Improving CX with conversational AI appeared first on Keatext.

What Marriott’s customer reviews tell us about CX in hospitality

Irene Serrano — Tue, 25 Aug 2020 18:12:45 +0000

When people think about the Marriott International hotel chain, they immediately think about consistency in luxury. That’s what customers expect and that’s what they get. Marriott’s secret to success isn’t a secret at all: it’s good CX that involves listening to customers, in person and through their digital footprint. Customer data tells the real story about everything that customers love and don’t love.

Keatext dove into over 4,000 customer reviews of Marriott hotels to discover the real reasons why customers book again and why they might stay elsewhere on their next trip. Applied to a voice of the customer program or any other CX initiative, these insights into customer opinions and choices represent an invaluable resource for tracing the customer journey and crafting CX strategies that work.

CX at Marriott, where technology enhances the customer journey

With over 7,000 properties and more than 30 distinctive brands worldwide, Marriott works to tap into localized industry trends in every country to attract new customers and grow, while also focusing on the global consistency that maintains loyal customers by exceeding their expectations, from online booking to check out.

Looking at CX specifically, it’s clear that Marriott International hasn’t shied away from exploring new technologies, as they have at every level of the company, while staying focused on the proven value of human interaction. In recent years, Marriott has explored both the physical and digital aspects of CX to create personalized experiences for its guests at every touchpoint, including guest rooms that accommodate Internet of Things smart devices and partnerships with tech companies whose tools enhance CX.

What Marriott’s data reveals about hospitality customer experience

Many companies already have an incredible amount of customer data at their fingertips — they just don’t always know how to use it. To prove this point, the Keatext’s team analyzed Marriott’s publicly available, unsolicited reviews from review sites where customers feel free to speak colloquially and in the language of their choice — in other words, much differently and candidly than they might when filling in a customer questionnaire or survey.

All insights start with data: Naming the sources for Marriott’s reviews

Marriott hotels reviewed: The New York Marriott Marquis, San Francisco Marriott Union Square, Chicago, Marriott Downtown Magnificent Mile and JW Marriott Austin.
Sources of reviews: TripAdvisor, Booking and Expedia.
Timeframe of reviews: 2016-2018
Number of reviews analyzed: 4,101
Number of separate comments analyzed: 10,245
Digging into the findings: How reviews reveal actionable insights

Positive sentiments, also known as praises, across channels: 5,327
Negative sentiments, also known as problems: 4,454
Number of customer suggestions found: 433
Number of questions from customers: 31
Main topics overall: Hotel, Room, Front desk, Service, Bed, Buffet breakfast, Bar, View, Bathroom and Restaurant

Breaking down these findings further by hotel, text analysis shows the specifics of every praise and problem. Each of these data-based insights has the potential to be addressed by CX strategists and reach across departments to address other concerns.

Out of 3,371 comments at The New York Marriott Marquis, 1,628 were praises and 1,576 were problems.

Among the praises: guests find the hotel clean, convenient and well-located; the rooms are considered spacious; the gym is well equipped and the view from the gym is appreciated.
Among the problems: a major point of contention is with the elevators, which customers say are too slow and often out of order; some customers note that the decor is dated and the food and drink expensive.

Out of 3,199 comments at San Francisco Marriott Union Square, 1,806 were praises and 1,289 were problems.

Among the praises: guests find staff friendly and the service good; they also praise the food quality and portion sizes.
Among the problems: Guests find the bathrooms small and parking very expensive; they consider noise pollution from the street to be heavy; they didn’t appreciate the union strike since services guest services were limited during that period; guests are not happy that the executive lounge is closed on the weekends.

Out of 3,558 comments at the Chicago Marriott Downtown Magnificent Mile, 1,858 were praises and 1,566 were problems.

Among the praises: guests enjoy the location and the view; they appreciate that the hotel is located so close to many restaurants; the breakfast buffet has a good selection; the food served at conferences is excellent.
Among the problems: some guests said that desk chairs in rooms are not easily movable and cannot be raised or lowered, and suggested that proper desk chairs for working would be appreciated; guests also would like to see refrigerators and microwaves in all rooms; they noted that electrical outlets are lacking and badly placed.

Out of 1,928 comments at JW Marriott Austin, 1,259 were praises and 605 were problems.

Among the praises: guests said the staff is friendly and helpful; they enjoy the spa and the gym; they appreciate the warm water and the view from the pool.
Among the problems: some guests considered the massage prices at the spa expensive; some noted that the pool space is too limited for the number of guests using it; and think the wifi is too expensive or should be free.

How Marriott can take the next steps to CX success

Considering the scope of this customer data and what it reveals, what could Marriott do next to address every insight? Customer reviews like the ones Keatext analyzed above reveal more than any star-based review system could. These reviews, understood as qualitative data that is too often gathered but not used by companies, help CX leaders and other departments understand the true reasoning behind customers’ opinions and sentiments.

Seen through the lens of CX, these insights pinpoint the exact steps in the customer journey that need more attention or need to be changed. In that way, Keatext’s insights can go far to address guest satisfaction, customer retention and even how to gain new customers along the way.

The post What Marriott’s customer reviews tell us about CX in hospitality appeared first on Keatext.

How Breckenridge Grand Vacations improved their customer experience with text analytics

Irene Serrano — Tue, 26 Mar 2019 16:59:19 +0000

In many industries, PR and sentiment analysis are designed to support crisis management and to react nimbly in the face of negative feedback or press. But when it comes to the service and hospitality industries, brands truly gain a competitive advantage by using sentiment analysis not only to nip problems in the bud, but to proactively find new opportunities to further improve and simplify every step of the customer journey.

That’s precisely how Breckenridge Grand Vacations (BGV) chose to leverage Keatext’s text analytics functionalities—and they saw promising results. Find out how they gained valuable visibility across unstructured feedback channels, simplified internal communication and decision-making, and boosted customer happiness by turning feedback into action.

Spotting opportunities

If you’ll forgive us the cliché, it’s often said that a good offense is the best defense. With today’s ever-expanding big data and AI capabilities, the secret to finding those charming offensive tactics is often nestled within a company’s multiple channels of complex, unstructured data. Those with the right tools and curiosity to contextualize and leverage that data find themselves tapping into a goldmine of invaluable insights into their customers’ wants and needs. That, of course, is where they gain their competitive edge. It’s also where Rick Tramontana, BGV’s Director of Owner Relations, saw big potential for exciting new ideas and opportunities.

“We really saw Keatext as an opportunity to make what was already good that much better, and to take it to the next level.”

“Ultimately, our model is to give our owners the absolute best experience possible. That way, the next time they’re up at a sales table, they’re more likely to buy more from us,” Rick said. “So we really saw Keatext as an opportunity to make what was already good that much better, and to take it to the next level.”

Case in point: through the data, BGV could proactively alter policies to adapt and optimize them for a whole range of reservation types and sizes—and do so before complaints come in.

By combing through text analytics feedback, Quality Assurance Analyst Katy Bath was able to pick up on a few minor friction points that were needlessly adding complexity to cancellations and reschedulings. With the right data in hand to move up the decision-making ladder, Katy and her team were able to quickly implement small changes to their cancellation policies, then track customer satisfaction in real time as it rose in response.

Finding the right tools

For the Breckenridge Grand Vacations team to achieve their goals, simple implementation with existing tools was essential to internal buy-in. The company was also looking for a tool that would allow them to track sentiment across extensive periods of time, helping them follow the impact of key campaigns, milestones or corporate decisions. But most importantly, says Rick, they wanted to be able to compare and contrast sentiment analysis across different surveys, channels and user segments.

By combining Keatext’s text analytics with existing customer management infrastructures, the team was able to eliminate information silos and provide essential visibility across multiple channels at both the macro and micro level.

By combining Keatext’s text analytics with existing customer management infrastructures, the team was able to eliminate information silos and provide essential visibility across multiple channels at both the macro and micro level.

“Keatext helped us identify smaller trends, sure, but also larger trends that could be improved either at the location level, across one of our departments, or even at an entire company scale,” said Rick.

With Keatext’s unique text analytics algorithms, his team was able to create custom dashboards that reached across channels to track and contrast their most important metrics, often unveiling surprising insights along the way.

The verdict is in

With text analytics now fully integrated into the quality assurance process, Katy says it’s become a central part of the team’s tracking and improvement methodology.

“I love that we can have almost instant feedback and be able to share it with our teams,” Katy said. “To be able to measure the sentiment of some of these clients and see real impact—see the areas where things are going great, and also where there’s room for improvement. I really love being able to track that on a day to day basis.”

Beyond serving as an invaluable set of tools and technologies, text analytics also ushers in a new mindset and approach to customer feedback. Inherent in sentiment analysis is a focus on proactive learning, active conversations with customers, and a careful balance of both granular and global considerations. Great, powerful brands have elegantly walked this line for years—often by building their own internal algorithms and technology. But as consumers agree to share more and more of their data and feedback, they’ll increasingly expect that level of responsiveness and agility from companies of all sizes. That means those who learn to harness the power of AI and text analytics early are likely to be rewarded with growth, engagement and loyalty.

The post How Breckenridge Grand Vacations improved their customer experience with text analytics appeared first on Keatext.

Answering FAQs about unstructured data and text analytics

Irene Serrano — Fri, 18 Jan 2019 19:01:55 +0000

Customer feedback isn’t always pretty. In form and in content, text-based feedback falls into the vast realm of “unstructured data,” packed with valuable information but difficult if not impossible for traditional data analysis to accurately parse. Thankfully, we’ve got AI tech like Keatext. Here, we’ll dive into what unstructured data comprises and why AI machine learning and deep learning is the answer to unravelling your customer feedback into useful data.

What is unstructured data?

In the context of Customer Relationship Management (CRM) and Customer Experience Management (CEM or CXM), unstructured data refers to customer feedback that does not “tick the box,” that is, it doesn’t fit a predefined model of organization. Usually presented in text form, unstructured data constitutes qualitative data, including emails, phone transcripts, comments, reviews and social media posts, as well as open-ended survey responses.

How does Keatext help us understand unstructured data?

The way people talk and write when casually expressing themselves would make few high school language teachers proud. When emotions are involved, we tend to be even less coherent and often end up sending mixed messages. This kind of feedback can be hard for industry professionals to understand, let alone for systems that rely on a finite list of predefined keywords and rules. Keatext’s AI technology is far more flexible. Unconcerned with whether or not the topics appear in a list, Keatext rapidly sifts through the text, including written interjections such as “um” and “aha” and other throwaway parts of speech, to accurately identify what matters most to customers.

What is text analytics?

In the context of customer experience management (CEM or CXM), text analytics is a software tool for extracting meaning from written customer feedback (unstructured data) so that relevant action can be taken. It can be used to analyze solicited market research (survey) answers or unsolicited feedback such as emails, reviews and social media posts.

Text analytics vs. text mining – is there a difference?

Text mining refers to the examination of natural language text, the term used for texts written by humans to express themselves. The purpose of text mining is to identify key concepts, themes and patterns in the text, which it capably does, though without taking the original context of the writing into account. By manipulating the concepts, themes and patterns that were identified, text analytics helps create understanding in decision-makers, which leads to solving business problems.

What is machine learning?

Machine learning is a type of artificial intelligence (AI). It refers to the ability of software to “learn by itself.” This means an application takes data that is fed into it, looks for patterns and discerns probability, then makes statements, decisions or predictions based on its findings. With feedback, it becomes more accurate over time.

What is deep learning?

Deep learning is a sub-field of machine learning. It refers to the ability of the algorithm to learn features and tasks directly from data. In other words, the algorithm learns by looking for structure in the data it receives. The more data it receives, the better it learns and the more accurate it becomes.

What type of machine learning does Keatext text analytics use?

Keatext uses unsupervised machine learning. This enables it to uncover unexpected outcomes. When supervised machine learning is used, the algorithm trains against known outcome examples and, in the words of one of Keatext’s clients, “finds what you’re looking for.” Keatext text analytics is not biased in this way. It finds what is there to be found or, as the same client said, “what you’re not looking for.”

The post Answering FAQs about unstructured data and text analytics appeared first on Keatext.

How BRP Turned 10 Years of Surveys into Insights [Case Study]

Irene Serrano — Wed, 21 Nov 2018 16:57:37 +0000

Bombardier Recreational Products (BRP) is a global leader in powersports vehicles and propulsion systems that has built a name for itself crafting beloved products like the ubiquitous Ski-Doo snowmobile and Sea-Doo watercraft. Yet in 2015, they faced a new challenge in understanding their customers’ wants and needs. They had a wealth of meaningful unstructured product feedback data but no clear way of turning it into insight. That’s when they sought help from Keatext.

Spearheading their efforts, as BRP’s Global Customer Advocate, was Myshka Sansoin, who was hired in 2014 to lead the company’s VoC program. Her principle mandate: to help BRP understand their customers’ needs and satisfy them. No pressure, right? Should be straightforward. Though perhaps not, when you have nearly 10 years of surveys and contact center transcripts to make sense of. To get there, Myshka required a text analytics solution that would meet three main challenges.

Challenge 1: Unstructured, siloed product feedback

As Myshka searched for the right solution, product feedback was pouring in directly from consumers and indirectly from dealerships through a wide range of channels: social media, forums, Salesforce case requests, surveys and call centers. Moreover, a better part of the feedback was the result of open-ended questions, which meant customers were often communicating through free-form language, making it even more difficult to aggregate and contextualize. At the time, BRP’s approach was to go through the feedback manually, which was painfully time and labor-intensive: “We didn’t have a method of putting it together to get a 360-degree view of what our customers were telling us. We needed an easier way to leverage and understand the feedback, because we knew it was so rich.”

Myshka was looking for a system to quickly aggregate data from different channels and accurately uncover what customers thought, felt and desired. Only then would she have enough tangible data – in other words, proof – to influence internal stakeholders to make agile, customer-centric decisions that would fulfill the company’s needs.

Challenge 2: Multilingual sentiment analysis

While numerical data was obviously key to providing accurate insights and context for the company’s decision-makers, Myshka knew they needed more than word counts – a unique tool was required to help them understand the sentiment and emotions behind the words. Was “better” being used to say people expected better, or that they had never received better customer support? What’s more, given BRP’s global customer base, the tool would need to understand and aggregate linguistic nuances in the two languages most of the feedback was expressed in: French and English.

Challenge 3: Frictionless implementation

By removing the manual work and biases from BRP’s VoC research, the company’s best people would be freed up to proactively find new ways of improving their products, building engagement and addressing customer dissatisfaction. But there was no point in making this team available only to put them through months of grueling implementation and testing before seeing solid results. The solution needed to be effortless and quick, while providing crucial independence from the company’s IT and development departments.

The Solution: Keatext

With a clear understanding of everything the text analytics solution needed to provide, Myshka knew right away that Keatext was up to the task: “When I looked at different solutions, I quickly concluded that Keatext was the best in the business. Not only are they experts in artificial intelligence and feedback interpretation, but their collaborative approach and flexible structure was really impressive. Their adaptability was important to us, and they readily adjusted to our requirements when possible. Even now when we need support, they’re immediately responsive. For us, that’s essential.”

That easy collaboration gradually permeated BRP’s relationship with its customers as the company integrated essential Keatext metrics into their problem-solving and decision-making. Now that they had accessible, accurate insights into user sentiment, BRP took it to the next level – fostering transparency and collaboration not only between different teams and decision-makers within the company, but between the company and its biggest stakeholders: its customers.

To tackle the diversity and intricacies of BRP’s unstructured feedback, Keatext’s complex Machine Learning algorithm was specifically trained to aggregate and contextualize multilingual customer experience data and transform it into easy-to-understand reports. This allowed Myshka and her team to take the pulse of their customer base at a glance.

Myshka’s favorite features for product feedback analysis

Text categories: Keatext’s algorithm divides your feedback into four categories: problems, praises, questions and suggestions. Her favorite? The suggestions category, which offers up solutions and ideas for improvements directly from customers.
Sentiment score: After uploading your data to Keatext, you quickly get a sentiment score that functions as a reliable KPI and tracks overall customer sentiment. The cherry on top? For a more granular sense of what’s working and what’s not, you can filter the sentiment score according to topic or product.

As for the implementation process, Myshka describes it in one word: seamless. All she needed to do to gain insight was click on a link and create a username and password. No informations system implementation, no devops requirements. It was that easy.

The Result: A corporate shift

From a logistical standpoint, BRP’s work with Keatext allowed it to gain agility, act quickly and prioritize solutions based on customer needs and wants in their product feedback. Fundamentally, says Myshka, it helped the company include customer feedback in its daily operations: “The biggest improvement Keatext brought to us is facilitating or even participating in our culture change. BRP wants the customer to be at the center of everything it does. Having an easy way to understand and share what the customer wants helps us accomplish this.”

As the company continues to grow and adapt to shifting industry trends, BRP can rest assured they have a finger firmly planted on the market’s pulse – allowing them to continuously develop solutions, launch products that customers not only need, but love.

The post How BRP Turned 10 Years of Surveys into Insights [Case Study] appeared first on Keatext.