Why high-value shoppers still need a salesperson to buy online

Ecommerce has spent twenty years removing friction. Search, filters, reviews, saved cards, one-tap checkout. A five hundred rupee purchase is now a transaction that takes under a minute, and most of the industry's best work went into making that true.

Apply the same machinery to a two lakh rupee purchase and something different happens. Every step gets faster except the one the shopper actually needs help with. They are not short of information by the time they reach your product page. They are short of confidence.

High-value purchases are decisions, not transactions

At five hundred rupees the cost of being wrong is mild annoyance. The shopper buys, and if it is not right they return it. Buying is the cheapest way to find out, so they buy.

At two lakh they cannot run that experiment. Returning a sofa means a truck. Returning a ring may not be possible at all if it was sized or made to order. Delivery takes weeks rather than days, and the purchase usually sits inside something larger: a wedding, a new home, a gift that has to land.

So the shopper stops behaving like a buyer and starts behaving like a researcher. They open tabs. They ask a relative. They read the same three reviews twice. And in a lot of categories they finish that research online and then go to a store to actually decide, which means the website paid for the visit and the showroom took the credit.

A product page answers questions, but not every question

A product page is a broadcast. It answers the questions somebody anticipated while building it, which are the general ones: what it is, what it is made of, what it costs, what other people thought.

The questions that stop a sale are not general. They are about this shopper. Their room, their hand, their budget, their occasion. Nobody could have anticipated them, because they are not really about the product at all.

Static content scales. Judgement does not, which is why it is still standing in your store.

Three hundred and sixty degree views, swatch images and dimension tables are all genuine improvements, and they all answer the same question: what is this. None of them answers the question the shopper is actually stuck on, which is whether it is right for them.

The questions that show up just before checkout

Jewellery makes the pattern easy to see, because the gap between a photograph and the real thing is at its widest. A shopper who has read everything on the page still arrives at these:

• How does it actually look in person, rather than under studio lighting?
• Is this the right size, and what happens if it turns out not to be?
• How will it look worn, on someone like me?
• Is there another design that would suit me better?
• How does the diamond really look at this price?
• Which of these two would you recommend, for my budget?

Notice what these have in common. Not one of them can be answered by publishing more content. Each needs somebody to look, form a view and say it out loud. They are sales conversations, not information retrieval.

The category changes the question and not the shape of it. In furniture it is whether the fabric is really that colour and whether the piece will turn the corner of a staircase. In appliances it is which of five models suits this room. The shopper is always asking somebody to close the gap between what is published and what applies to them.

What a person adds that a page cannot

• Trust. Someone answering an awkward question in real time is a different kind of evidence from a badge on a page, because it cannot be faked in advance.
• Personalisation. A recommendation that accounts for a budget and a room is worth more than a list sorted by popularity.
• Comparison. Two products side by side, with someone explaining the one difference that actually matters, is something no product page is structured to do.
• Reassurance. Most of the hesitation before a large purchase is not a missing fact. It is wanting a second opinion from someone who has seen a hundred of these.
• Objections. An unanswered objection does not announce itself. It leaves.

The missing layer in the ecommerce funnel

Most funnel work optimises the steps you can see. The step that costs the most in high-value categories is the one that leaves no trace.

That exit does not arrive in a report labelled as an unanswered question. It arrives as a bounce, or an abandoned cart, which look like price problems. So they get treated as price problems, with a discount, and the margin goes to shoppers who were never hesitating about money in the first place.

Bringing the salesperson to the website

Assisted selling is the practice of putting a real person into the online journey at the point the shopper needs one. In its current form that usually means a one to one video consultation started from the product page, where an expert can show the item, answer the question and stay until the shopper has decided.

It is worth being precise about what this is not. It is not a call centre, where the shopper leaves the site and explains themselves from the beginning. It is not live chat, which handles factual questions well and judgement badly. And it is not a video meeting link, which asks a stranger to schedule something with a business they have not decided to trust yet.

This is the layer Winklook builds. A shopper taps for help on the page they are already reading, reaches an available expert, and the conversation happens where the product is.

What a modern assisted sale looks like

1. 1.A shopper is on a product page and wants to check something before committing.
2. 2.They ask for help, and reach an expert who is available now or book a time that suits them.
3. 3.The expert arrives already knowing which product is on screen, so the conversation starts at the question rather than at the beginning.
4. 4.They show the product, compare it against the obvious alternative and answer what was actually being asked.
5. 5.The shopper's details and what they were interested in are captured, whether or not they buy today.
6. 6.If they do not buy today, somebody follows up with the context of the conversation rather than a generic reminder.

What ecommerce teams should measure

Assisted selling stops being an experiment and starts being a channel at the point it is measured like one. The numbers worth putting on a dashboard from the first week:

• Assisted conversations, so the volume is visible at all.
• Connection rate, meaning how often a shopper who asks for help actually reaches somebody.
• Leads generated, including the conversations that did not end in a sale that day.
• Assisted conversion, compared against the same pages unassisted.
• Revenue generated, as far as your own ecommerce data allows it to be attributed.
• Response time, which is the number that quietly determines all of the others.
• Sales agent activity, so the online team is managed the way the floor team already is.

Without these it stays a feature somebody liked, and features that nobody can put a number against are the first thing cut in a budget review.

Available, not unavoidable

None of this means every high-value shopper wants a salesperson. Many would rather decide alone, and a store that interrupts them is a worse store, not a more helpful one. Assistance that cannot be ignored stops being assistance.

The opportunity is not to choose between digital convenience and human help. It is to make human help available at the moment it becomes valuable, and invisible for the rest of the visit. For a considered purchase, that moment arrives more often than most product pages are built to notice.