What do customer reviews and return reasons reveal that sales dashboards miss?

An e-commerce dashboard can tell a retailer which products are selling, which categories are growing and where revenue is coming from. But it may not explain why customers are disappointed after placing an order.

A product can generate strong sales and still leave customers unhappy. Its rating might be high, but reviews may repeatedly mention poor material quality, misleading photographs or a fit that differs from expectations. Return data may show a problem, while customer feedback explains what is causing it.

For e-commerce businesses, these signals matter because a completed purchase is not necessarily evidence of a successful customer experience.

Research on online reviews and product returns shows why businesses need to look beyond sales performance to understand the quality of the purchase decision itself.

1. Why can a product sell well and still disappoint customers?

Sales figures capture what customers decided to buy. They do not necessarily tell retailers whether the product met the expectations that influenced that decision.

Consider a simple example: an online marketplace sells a popular office chair. The product page has attractive images, a competitive price and a high average rating. Orders increase.

But after delivery, customers begin mentioning that the chair feels smaller than expected, the cushioning is thinner than it appears online, or assembly is more difficult than anticipated.

The sales dashboard may continue to show strong demand. The reviews reveal a different part of the story: customers are buying the product, but some are not receiving the experience they expected.

Research supports the importance of this distinction. A 2018 study published in Information Systems Research, The Impact of Online Product Reviews on Product Returns, analysed transaction-level data from a multichannel, multibrand North American specialty retailer over two years.

The researchers found that the availability of more reviews and more helpful reviews was associated with fewer product returns, after accounting for customer, product and other contextual factors. They also found that products displayed with an inflated average rating were returned more often.

The implication is that reviews do more than influence whether a customer clicks the buy button. They can also help customers decide whether a product is right for them.

For retailers, the lesson is straightforward: sales performance and post-purchase satisfaction need to be examined together.

2. What can return reasons tell retailers that return rates cannot?

A return rate identifies the scale of a problem. Return reasons help explain its nature.

Suppose two products have the same return rate. One is frequently returned because customers ordered the wrong size. The other is returned because its material feels different from what the product page suggests.

The numbers may look similar, but the solutions are different.

The first product may need clearer sizing information or a better fit guide. The second may need more accurate photography, material descriptions or quality checks.

A 2025 study published in the International Journal of Production Economics, Understanding and predicting online product return behavior: An interpretable machine learning approach, examined how customer reviews can help identify drivers of online product returns.

The researchers identified product attributes such as physical appearance and durability, alongside factors including packaging and service, as relevant to return behaviour. They also found that the impact of these factors varies across product categories.

This matters because a single return-rate metric can combine several very different problems.

Businesses should therefore classify return reasons into useful categories, such as:

• Product mismatch: The item differs from its description, photographs or stated specifications.
• Quality issues: Customers report defects, damage or concerns about durability.
• Fit and usability: The product does not fit, function or perform as expected.
• Delivery and packaging: The item arrives late, damaged or inadequately packaged.
• Post-purchase service: Customers experience difficulty with returns, refunds or support.

These categories are a practical starting point, not a universal classification established by the study. Retailers should adapt them to their products and customer language.

The objective is to move from knowing that a product is being returned to understanding what needs to change.

3. Can customer reviews reveal problems customers do not select as return reasons?

Yes. A customer might select "item not as expected" from a return menu, but that label does not explain what went wrong.

Was the colour different from the photograph? Was the material thinner than expected? Did the product lack a feature the customer assumed it would have?

A predefined return reason captures a category. The customer's written review can provide the detail needed to investigate the underlying problem.

A 2024 open-access study, Shaping the causes of product returns: topic modeling on online customer reviews, explored how natural language processing and topic-modelling techniques can identify return causes across product categories.

The researchers found that review-based analysis could reveal category-specific causes that generic return-reason menus may fail to capture. For example, apparel and footwear have different potential problems from electronics or home products.

This matters particularly for online marketplaces that sell products from multiple brands and sellers. A generic return menu might group several distinct issues together, making it difficult to identify which products, suppliers or descriptions need attention.

Consider these two reviews:

• "Not as expected."
• "The fabric is much thinner than shown in the pictures, and the lining is missing."

The first tells the retailer that expectations were not met. The second points towards specific issues that the product team can investigate.

Retailers can use review analysis to group recurring phrases and themes, then compare them with structured return reasons. The result is a more detailed view of customer dissatisfaction.

However, reviews are not a complete record of customer experience. Some customers never leave feedback, and those who do may not represent all buyers. Review themes should therefore be compared with return records, product inspections and customer support data.

4. What do customer reviews reveal about the expectations created by product pages?

Product descriptions, photographs, ratings and reviews all help customers form an idea of what they will receive.

When the delivered product differs from that expectation, the problem may begin before the order is placed.

A 2016 study published in the Journal of Retailing, To Keep or Not to Keep: Effects of Online Customer Reviews on Product Returns, analysed multi-year data from a major European online retailer covering electronics and furniture.

The researchers found that unusually positive reviews could increase both purchases and returns. One possible explanation was that overly positive reviews created expectations that the actual product did not meet. The effect was stronger for inexperienced buyers and cheaper products in the study.

This creates an important challenge for retailers. Strong ratings can help a product sell, but the goal should not be to maximise positive impressions at any cost.

Instead, businesses should examine whether customers' expectations match the delivered experience.

For example, if reviews repeatedly mention that a product is smaller than expected, the retailer could investigate whether dimensions are sufficiently prominent on the product page. If customers describe a colour differently from its listing, product photography and colour descriptions may need attention.

These are possible responses to investigate, not guaranteed solutions. The underlying issue could also involve manufacturing variations or differences between product batches.

The useful question is: what are customers expecting when they buy this product, and where does the actual experience differ?

5. What can return data reveal about the experience after a purchase?

Not every return is caused by a product defect or misleading description. The return process itself can become part of the problem.

Customers may struggle to understand eligibility rules, arrange a pickup, track a refund or get a clear response from support. These experiences can affect how they view the retailer, even when the original product issue is relatively minor.

The National Retail Federation's 2025 Retail Returns Landscape, published on 15 October 2025 in partnership with Happy Returns, projected that 19.3% of online sales in the United States would be returned in 2025. It also reported that 82% of surveyed consumers considered free returns an important factor when shopping online.

These are US-specific estimates and survey findings, not figures that should automatically be applied to Indian e-commerce. They nevertheless illustrate why returns deserve attention as part of the overall shopping experience.

Retailers should examine service-related feedback alongside product-related feedback. Comments about delayed refunds, unclear return instructions or repeated follow-ups can indicate friction that sales and inventory reports will not explain.

A high return rate may point towards product issues. A high volume of complaints about refunds may indicate an operational problem. If both appear together, the business may need to investigate the full journey rather than treating the return as an isolated transaction.

6. How should businesses connect reviews, return reasons and sales data?

The greatest value comes when these sources are examined together rather than in separate reports.

Imagine that sales for a pair of running shoes are growing, but return rates are rising too. Reviews repeatedly mention that the shoes run narrower than expected.

The sales dashboard shows demand. The return records show a growing issue. The reviews suggest a possible explanation.

The retailer could then investigate sizing guidance, product specifications and differences between product batches. After making an appropriate change, it could monitor subsequent returns and customer feedback to see whether the pattern improves.

This is more useful than treating sales growth as the only sign of success.

It also avoids assuming that every negative review signals a product failure. Some customers may have different preferences, misunderstand instructions or experience isolated issues. Patterns across multiple sources provide a stronger basis for decisions than individual comments alone.

Want to measure what happens after the click or the order? The CPaaS KPI Handbook covers 50+ WhatsApp, SMS and RCS metrics, with formulas and worked examples.

7. How can e-commerce teams turn these signals into action?

Collecting customer feedback is only useful if the business acts on what it learns. A practical approach is to build a regular review process around recurring problems.

Start with the products showing unusual patterns. Look for products with rising return rates, repeated complaints or a growing gap between high sales and negative post-purchase feedback.

Identify recurring themes. Group reviews and return reasons by issues such as product quality, sizing, missing features, delivery damage or refund delays. Review the original comments before deciding what they mean.

Assign the problem to the right team. Product-description issues may need attention from the content team. Quality complaints may require supplier or product-team investigation. Refund delays may need operational changes.

Measure whether the response worked. Track the relevant return reasons, review themes, customer complaints and sales performance after making a change. Compare similar periods and account for changes in product mix, promotions and demand.

For larger catalogues, text-analysis tools can help identify recurring themes across thousands of reviews. Human review remains important, especially when customer comments are ambiguous or automated classifications could misinterpret context.

The aim is not to eliminate every return or negative review. Returns can be legitimate, and critical feedback can help customers make better decisions. The aim is to understand which problems are preventable and where a change would benefit both customers and the business.

Want to keep customers informed after delivery and returns? Talk to the Helo.ai team about WhatsApp and SMS updates for your store.

The takeaway: sales tell you what happened. Customer feedback helps explain why.

Sales dashboards remain essential for tracking demand, revenue and product performance. But they offer only one view of the customer journey.

Reviews can expose gaps between product promises and actual experiences. Return reasons can identify recurring issues. Support records can reveal where customers struggle to resolve them.

When businesses connect these signals, they can make better decisions about product information, quality, fulfilment and customer service.

The opportunity is not simply to reduce returns or improve ratings. It is to understand what customers are telling the business, identify the problems that matter most and use that evidence to improve the next purchase experience.

A useful question for every e-commerce team is this: which products are performing well on the sales dashboard but telling a different story in customer feedback?

Frequently asked questions

Why do customers return products bought online?

Studies point to product attributes such as durability, packaging and customer service, and to expectations set by reviews and ratings. The effect varies by product category.

How can customer reviews help reduce product returns?

A 2018 study found that more reviews, and more helpful reviews, were associated with fewer returns. Products shown with an inflated average rating were returned more often.

Why combine sales data with return and review data?

Sales show demand, return reasons show what is going wrong and reviews explain why. Together they help teams find fixable problems such as unclear sizing or misleading photos.