What Customers Say About You vs What They Actually Do

What your customers say about you is every opinion they publish or report: star ratings, written reviews, survey answers, support messages, social posts and comments. You find it by reading your own reviews and tickets, by monitoring the rest of the web with tools such as Mention, Tagboard and Google Alerts, and by asking directly with on-site and post-purchase surveys.

The part most teams skip is what to do with it afterwards. What customers say is one kind of data, and what customers do on your site is another. They come from different places, they disagree more often than you would expect, and neither one is a lie. This article covers where to find what customers say, how to read it honestly, and how to confirm it against behavior before you rebuild a page around a comment.

What "what customers say about you" actually means

It arrives in three forms, and they are not interchangeable.

• Solicited feedback. You asked: surveys, NPS, review requests, interviews. You control the question, so you can aim it at a specific page or moment.
• Unsolicited feedback. They chose to speak: public reviews, social posts, forum threads, blog comments. You do not control the question, which is exactly why it surfaces problems your survey never thought to ask about.
• Service feedback. They needed help: tickets, chats, returns, complaints. This is the most concrete of the three, because every item is attached to a real order that went wrong.

All three describe experience. None of them measures it.

Say versus do: the gap you have to plan for

The gap is ordinary and predictable. Shoppers say price is the reason they left, then buy the more expensive bundle. They say they want more choice, then abandon a 40-item category page. They rate a checkout "fine" in a survey and drop out of it twice the same week. People report the reason that sounds sensible, not the reason that operated.

The practical consequence: never treat a survey answer as a finding. Treat it as a candidate explanation that has to earn its place.

Why it matters to know what customers say about you

Think about how you buy. You ask a friend which restaurant is good. You read the reviews of a television before Black Friday. Your customers do the same, and they do it about you. According to BrightLocal's Local Consumer Review Survey 2026, 97 percent of consumers read reviews for local businesses, and the share who say they always read them rose from 29 percent to 41 percent in a year. That survey covers local businesses rather than all of eCommerce, so read it as a direction rather than a benchmark for your store, but the direction is not in doubt.

You learn more about your own business

Feedback tells you what is going wrong, where, and for whom, and it also tells you what already works so you stop redesigning it. Your job as a seller is to meet demand you did not invent, and the fastest description of that demand comes from the people who paid.

You learn more about your competitors

Your customers shop your competitors too, and they will tell you about it. When someone explains why they left another store, read it closely: why did they stop buying there, and what made them choose you? Their competitor reviews are a free audit of what a rival offers that you do not.

You can anticipate what customers will want next

Praise ages. If customers currently love your next-day delivery, and two competitors now offer the same, the compliment is about to become the baseline. Take what makes people happy today and improve it before the market forces you to.

You prevent bad experiences from spreading

No store is perfect, and a single bad review is not a crisis. A repeated bad review is, because it is public, permanent and read by the majority of buyers before they decide. Reading feedback weekly turns a pattern into a fix while it is still small.

Where to find what customers are saying

1. Read your own reviews and tickets first

   Sort by the lowest ratings and by the reviews that contain a suggestion, then list every complaint that appears more than twice. This costs nothing, it is specific to your catalog, and it usually fills a quarter of your roadmap.
2. Monitor mentions across the web

   Mention tracks who talks about you across social platforms, blogs, review sites and press, and it does the same for competitors and category terms. Use it to catch conversation that never reaches your own properties.
3. Track your hashtags around campaigns and events

   If you run a branded hashtag, or one for an event or promotion, Tagboard collects and displays the posts that use it. This is the fastest read on how a campaign landed while it is still running.
4. Set up Google Alerts for brand and product terms

   Alerts for your brand, your key products and common misspellings deliver new web results to your inbox. It is crude compared with a monitoring platform, it is free, and it catches the article nobody told you about.
5. Ask directly, at the right moment

   Listening only collects the opinions people volunteered. Surveys in Omniconvert Explore let you ask a specific question at a specific step, on-site during the visit or after purchase, so you learn about the pages nobody writes reviews about.
6. Put it all in one list, grouped by theme

   Feedback in five tools is feedback in no tool. Once a month, merge everything into a single list grouped by theme, with a count next to each theme. Frequency, not volume of words, is what tells you where to start.

How to check what customers say against what they do

This is the step that converts feedback into money. A comment on its own is a story. The same comment, written as a hypothesis and tested against the live site, is a decision you can defend.

1. Write the complaint as a falsifiable hypothesis

   "Shoppers abandon the cart because the delivery cost appears too late" can be wrong. "Customers want a better experience" cannot, so it is useless.
2. Change one thing that removes the obstacle

   Show the delivery cost on the product page. Keep the rest of the page identical, or you will not know which change did the work.
3. Run it as an A/B test with a money metric

   Conversion rate, revenue per visitor or completed checkouts. Satisfaction with the new page is not evidence that the new page sells more.
4. Read the result by segment

   A change often helps mobile visitors and does nothing on desktop, or helps new customers and annoys returning ones. A flat average hides both.
5. Feed the answer back into the next question

   A losing test is not a wasted test. It removes one explanation, which narrows the next survey question to something sharper.

Deciding what you want customers to say about you

The trick is not to discourage reviews or silence criticism. It is to choose the praise you intend to deserve. For an eCommerce business, the shortlist usually looks like this:

• They have the best products.
• They always keep me informed.
• They are problem solvers.
• They respect me.
• I trust them.

Pick three. For each one, name the process that produces it, the person who owns it and the moment the customer feels it. "They always keep me informed" is a shipping notification schedule and a support reply time, not a value on an About page. Then ask for the review immediately after that moment, because a customer who has just been kept informed writes a different review than one asked six weeks later.

NPS and how to choose NPS software

Word of mouth still sells, so knowing whether customers would recommend you is worth measuring. Market research metrics are often complex; NPS is not, which is why it survives. A falling score is an early warning that the experience is slipping. A rising one is evidence that a customer-centric strategy is working. Because it is measured the same way every time, it is comparable from one quarter to the next.

The score alone is thin. Its usefulness comes from the free text underneath it and from segmentation: an NPS of 40 that hides a 10 among first-time buyers is a very different business problem from a flat 40 everywhere.

When comparing tools, check for:

• A true 0 to 10 scale, so your score is comparable with everyone else's.
• Flexible triggers, so you can ask after delivery or after support contact, not on every page view.
• Follow-up questions, so respondents explain the number.
• Segmentation, by traffic source, device, product, order history and customer value.
• Readable reporting, showing the trend and the verbatim answers side by side.
• A path to action, ideally a survey tool inside the same platform as your testing tool, so an answer becomes an experiment without an export.

If retention is the reason you are asking, connect the answers to customer value. Nexus by Omniconvert segments customers by recency, frequency and monetary value, so you can read what your highest-value customers say separately from everyone else's opinion.

Frequently Asked Questions

Andrada Vonhaz

My journey in marketing and technology has always been fueled by a relentless passion for innovation. It all began in 2016 when I delved into the fundamental aspects of content and copywriting, alongside mastering social media dynamics, SEO, and PPC tactics. By combining these elements, I cultivated a deep understanding of marketing strategies, polished my project management skills, and became adept at communicating and monitoring marketing initiatives from inception to completion. My guiding principle is rooted in the belief that there’s always something new to learn and apply. It’s this mindset that drives me to continually evolve and innovate in the dynamic worlds of marketing and technology.