eCommerce Customer Service

The conversation that decides the sale

Online shoppers contact you at two moments: when they are close to buying and something is unclear, and when something has gone wrong. The first is the most commercially valuable conversation in retail and the most commonly under-resourced.

A pre-purchase question is a customer telling you exactly what is blocking their basket. Answer it in ninety seconds and you often make the sale. Answer it in four hours and they have bought elsewhere, and you never learn that the question existed.

RHI provides dedicated retail customer service teams staffed for both moments, with the response speed the pre-purchase window actually requires.

Pre-purchase, in detail

These contacts are short, high-volume, and resolvable in one exchange if the agent has the information. Typically: will this fit, will it arrive by a date, is it compatible, is it in stock in a size, what is the return policy really, can I change the delivery address after ordering, does the discount apply to this item.

Three things make the difference. Speed, because the customer is on the page now. Product knowledge, because "let me check" costs the moment. And permission to be definite - an agent who can only quote the policy page is less useful than the policy page.

Live chat carries most of this volume, which makes concurrency the design decision that matters - see live chat support. We set it per queue, and pre-purchase queues run at lower concurrency than order-status queues because the value of a fast, attentive reply is higher.

When something has gone wrong

The other half of the volume, and emotionally the harder half. A delivery that did not arrive, an item that is not what was pictured, a charge that looks wrong, a refund that has not appeared.

Customers arriving here are frequently already annoyed, often already looking at the returns page, and increasingly likely to post publicly if the response disappoints. Handling it well needs authority over the order rather than sympathy - see eCommerce support for how the operations side works, and returns and refunds for the reverse journey.

One thing worth resourcing deliberately: the recovery contact after a failure is resolved. A customer whose delivery went wrong and was then fixed properly is frequently more loyal than one whose order was uneventful, but only if someone closed the loop rather than leaving them to notice. That follow-up is cheap, rarely staffed, and it is where a bad experience converts into a retained customer.

Selling without becoming a sales team

Retail support agents are in a position to help customers buy, and there is a line worth drawing carefully.

Helpful: answering the blocking question definitively, suggesting the correct size when the customer described a fit problem, mentioning that the item they want is available in another colour, telling them a bundle costs less than the two items separately. That is service, and it converts.

Not helpful: pushing an upsell during a complaint, recommending a more expensive item that does not fit the stated need, or treating a support contact as a sales opportunity. It reads as opportunistic and it damages the contact that was going fine.

We train the first and score against the second. Where you want a defined recommendation prompt in specific contact types, that is agreed and trained rather than left to individual judgement.

Peak trading and the first-impression problem

Peak volume arrives with new customers attached. During a sale, a materially higher share of your contacts come from people who have never bought from you before and are forming their entire opinion of your service from this one interaction.

That is the argument for resourcing peak properly rather than accepting degraded service for a fortnight. A core dedicated team plus planned surge cover, with the surge team recruited and trained in advance, is what makes it work - see overflow support. Surge needs months of lead time, not weeks.

Languages, coverage and systems

Retail contact skews heavily to evenings and weekends, because that is when people shop. A weekday-office roster misses a large share of the pre-purchase window entirely. Coverage is built from shift units.

From India we staff English, Hindi, Hindi-English bilingual, Bengali, Nepali and Punjabi, with further Indian languages on request - see our languages page. For international markets we staff neutral international English.

Agents work in your commerce platform and help desk so they can see the order while discussing it. Our teams are experienced across Shopify, WooCommerce, Adobe Commerce, Amazon Seller Central, Zendesk, Freshdesk, Gorgias and Zoho Desk - competence, not partnerships. See technology.

Quality and reporting

Reported separately for pre-purchase and post-order, because they have different targets and blending them hides both. Pre-purchase: first response time, resolution in one exchange, and where you track it, contact-assisted conversion. Post-order: resolution time by reason, refund and replacement turnaround, repeat contact.

Across both: contact rate per order, satisfaction with free-text themes read rather than counted, and quality score. Interactions are sampled and scored against an agreed scorecard - see quality assurance. Figures published elsewhere on this site are historical or representative and campaign-dependent.

Getting started, and what it costs

Priced per dedicated agent per month with a five-agent minimum - published rates and what moves them.

Bring contact volume by hour and weekday, your split between pre-purchase and post-order if you have it, and your trading calendar. If you cannot split pre-purchase from post-order, that is worth measuring first - it usually turns out to be a larger share than expected, and it is the share with revenue attached. Talk to us.