Buying review software or building audience research in-house?

What to take away

Why audience research on reviews is different from ordinary customer research

Most customer research asks what people bought and why. Review research asks something narrower: what does a shopper need to see in a review before they trust it enough to buy?

That distinction matters for an English retailer because review behaviour is local. A shopper in Bristol checking a review widget on a product page is doing something different from a shopper reading a Trustpilot page after a delivery problem. The audience, the context and the trust signal differ.

Which? tracks consumer attitudes and trust across the UK through its Consumer Insight Tracker. The pattern it reports is consistent: trust in a claim depends heavily on who makes it and where it appears. That is the starting point for any review research programme.

What do you already know before you ask anyone anything?

Before booking interviews or writing a survey, mine the data already in your business. Most small retailers have more review intelligence than they realise.

Look at four sources: the text of existing reviews, the reasons customers give for returns, the questions support handles most often, and the search terms people use on your own site. Each produces language you can reuse in questions later.

Tag each item by theme as you work through it: delivery, quality, sizing, returns, service. A spreadsheet column is enough. You are hunting for the words customers repeat, not for a clean statistical sample.

The Office for National Statistics publishes home internet and social media usage data that shows how differently age groups use online channels. Use it to sanity-check assumptions about which customers read reviews on a phone versus a desktop.

Write down what you think you know as a list of assumptions, then mark each one as evidenced, guessed or unknown. The unknowns become your research questions.

Who is actually reading your reviews?

A persona is not a demographic label. For review research it describes how someone shops and what they need proved before they commit.

Four personas cover most English retailers: the cautious first-time buyer, the repeat buyer checking for consistency, the gift buyer who does not know the category, and the researcher who reads twenty reviews before buying. Start with those and add only when the data demands it.

Building these properly takes more than a workshop. A structured approach to commerce reviews buyer personas in England shows how to ground each persona in real review language rather than assumptions about age or income.

Keep the set small. Four personas that your team can name from memory will be used; twelve will be filed and forgotten.

How do you run customer interviews that produce usable review insights?

Interviews are the fastest way to hear the words customers use when they describe trust, doubt and proof. Six to ten conversations will surface most of the patterns.

Recruit from your own customer list, not a panel. You want people who have actually bought, returned or abandoned a basket in the last six months.

Aim for a mix: recent first-time buyers, repeat customers and people who returned an item. Each group describes proof differently.

Ask about the last purchase, not purchase behaviour in general. 'Walk me through the last time you checked a review before buying' produces specifics; 'do you read reviews' produces polite noise.

Record the exact phrases people use. Those phrases become your survey options, your review request wording and eventually your on-site copy.

A fuller method, including question design and how to avoid leading answers, is set out in this guide to commerce reviews customer interviews in England.

Which survey questions measure review trust without wasting respondents' time?

Surveys give you scale. Interviews tell you why; surveys tell you how many. Keep them short enough that people finish them on a phone.

A workable structure is ten to twelve questions. Three cover shopping habits, four cover where people read reviews, three cover what makes a review believable, and two are open questions about a recent review.

Avoid asking people to rate review importance on a scale of one to ten. It produces a number nobody can act on. Ask instead which of two specific statements they agree with more.

The choice of tool matters less than the question design, but it does affect completion rates and how you segment results. This comparison of commerce reviews survey tools in England covers the practical trade-offs for small teams.

For in-house survey work, Google Forms and Microsoft Forms are free with an existing account. SurveyMonkey and Typeform add logic, branding and better reporting. Published small-team plans for those two typically start around £25 per month as of 2026.

Run the survey quarterly rather than annually. Review behaviour shifts with platform changes and seasonal buying patterns, and a stale dataset is worse than a small current one.

What can competitor review research tell you that your own data cannot?

Your own reviews show what your customers think. Competitor reviews show what the wider market expects, including the complaints people make about products like yours.

Read the one-star and three-star reviews on competitor product pages. Five-star reviews confirm expectations; middle and low reviews reveal the doubts your own product page needs to answer.

Compare how competitors handle review display: how many reviews appear above the fold, whether photos are shown, whether negative reviews are published at all. The presence or absence of negative reviews changes how believable the positive ones look.

IAB UK maintains an insight archive covering digital advertising and the formats buyers respond to. That is useful context when you are deciding how much of your review content should be paid amplification rather than organic display.

A structured method for this work, including which competitors to pick and what to record, is covered in this guide to commerce reviews competitor research in England.

Should you buy review software or build audience research in-house?

Software earns its place once you have a repeatable process and enough volume that manual collection breaks down. Buying it first usually means paying for features you have not yet defined a need for.

Use manual collection while you are running interviews and the first survey. You will learn which parts of the process are genuinely painful, and that tells you what to look for.

When you do buy, judge tools on four things: how they collect reviews, how they moderate them, how they display them, and what data they export. The export matters more than most buyers expect. It determines whether you can analyse review text alongside your other research.

Ask any supplier two questions before you sign: can it export your full review text, and can it show reviews collected on other platforms? If the answer to either is no, keep looking.

If internal teams will be asked to generate content or advocacy as part of the programme, the CIPD's resources on employee engagement are worth reading. Staff advocacy does not happen because it was requested. It follows from how people are treated at work.

A numbered process you can run in one quarter

How do you turn findings into review content that performs?

The output of audience research should be specific changes, not a report. If nothing on your product page changes, the research has not finished.

Common outputs include rewriting review request emails in the customer's own language and adding the questions buyers ask most to your product page. Another is changing which reviews appear first, so the useful middle-ground reviews are visible.

Segment your review display by persona where the platform allows it. A gift buyer and a repeat buyer need different proof, and showing both the same three reviews wastes the research.

The Chartered Institute of Marketing's content hub carries marketing research and guidance that helps connect these findings to wider campaign planning, particularly when review content is reused across email and social channels.

Track one metric per change. Review request response rate, review count per product and conversion rate on pages with reviews are all reasonable, provided you pick them before you make the change.

Write the change into the product page before the next campaign, so that the review content and the page agree with each other.

What changes should you expect over the next year?

Review behaviour is not static. Platforms change how they verify purchases, regulators take more interest in how reviews are presented, and shopper habits shift with them.

For an English retailer, the practical planning question is which parts of your current approach depend on a platform rule that might change. Verified purchase badges, review gating and incentive disclosure are the three areas most likely to move.

Keep a short note of which platform rules your programme depends on, and diarise a check each quarter.

Our outlook on commerce reviews trends and outlook for England in 2027 sets out what to watch and how to build a review programme that survives those changes.

Build the research rhythm now so that when a rule or platform changes, you already know which customers it affects and how to ask them.

Common questions

How many customer interviews do I need before the findings are reliable?

Six to ten is usually enough for a single product category, provided the participants are recent buyers. The aim is to surface language and patterns, not to produce statistics. If the tenth interview repeats what the first six said, you have enough.

Can I run audience research on reviews without buying software?

Yes, for the first cycle. Spreadsheets, your email platform, a free survey tool such as Google Forms, and a research repository such as Airtable or Notion will cover interviews and an initial survey. Buy software once you can describe the specific process it needs to support and the data you need it to export.

How often should I repeat review audience research?

Quarterly works for most small English retailers. Review platforms and consumer expectations move faster than annually, but more frequent fieldwork exhausts your customer list and produces samples too small to compare.

What is the single most useful data source for this research?

The text of your existing reviews, including the negative ones. It is free, already written by buyers, and uses the exact language your future survey questions and product pages should use.