Good answers.
More confident decisions.

Interest does not always mean readiness to buy.

Syndigo’s 2025 study surveyed more than 8,500 consumers across the US, UK, France, Germany, Brazil and Mexico. Its finding describes shoppers who reported this experience; it does not mean 44% of all orders are lost for this reason.

Why a longer product page is not always the answer

Baymard’s usability research observed shoppers leaving a site to research an unanswered question, or abandoning the purchase. It also found that trying to cover every use case in a product description can make it too long to read. FAQs and product Q&A help extend the description.[3]

Our interpretation: make relevant information easier to get, as well as making the product page better.

“Will this work for me?”

PowerReviews surveyed 15,870 US consumers in December 2023. Respondents also cited understanding product performance (79%) as a reason for reading Q&A. The sample was 88% women, so it should not be treated as a representative cross-section of all shoppers.

Our interpretation: useful guidance connects product facts to the shopper’s situation. It may take a follow-up question to understand that situation.

More information can still leave parents unsure.

NIQ and Momcozy’s 2026 research surveyed 1,240 US female caregivers aged 18–45 in March 2026. Participants were pregnant with their first child or had children aged 0–24 months.[4]

These findings identify decision difficulty within this group. They do not establish that every parent has unanswered questions, that baby products create more questions than every other category, or that AI by itself improves conversion.

Our interpretation: clear, traceable product guidance matters. For suitability or safety questions, answers need to stay within approved product instructions and hand off when a qualified person is needed.

A bigger purchase can need more reassurance.

Northwestern University’s Spiegel Research Center reports a larger conversion effect from displaying reviews for higher-priced products than for lower-priced products: 380% versus 190% relative increases in the studied setting. The findings were first published in 2017 and revisited by the center in 2025.[5]

This is evidence about customer reviews. It is not a measurement of AI support, and it does not prove that AI’s impact always increases with price. We use it to inform our focus on purchase uncertainty, not as a GoNovus performance claim.

Our interpretation: for considered purchases, comparison, compatibility and use-case questions deserve particular attention.

What happens when a shopper gets an answer sooner?

A paper by Amazon researchers, accepted at ECIR 2024, evaluated an automated question-answering system with millions of e-commerce customers. Purchase rate was defined as the share of users asking about a product who bought it within one week.

The paper reports relative results, without full randomization details or absolute purchase rates. This is a result from that deployment, not an expected uplift for every store or a GoNovus result. We also exclude the larger figure for users who explicitly rated the answer helpful, because that selects a particular subset of shoppers.

Our interpretation: timely answers are a credible path to test. The impact for a store still needs to be measured in that store.

Resolution follows your business rules.

This is a GoNovus customer result, separate from the third-party pre-sale research above. Results vary with the workflows covered, connected information and the business’s escalation policies.

Agree on what “resolved” means.

During evaluation, define the requests in scope, the measurement period, and how human handoffs and reopened requests are counted. A reply alone is not enough to demonstrate resolution.