The cost of slow customer service in ecommerce hides well. Nobody sends an email saying "I bought elsewhere because you took nine hours to answer." The order simply never appears, and the shopper becomes a line in your analytics that says "abandoned."
For a small store, those invisible losses add up to a real number. This post works through a hypothetical example so you can do the same sums with your own figures. The numbers are illustrative, not drawn from any particular store.
The first cost: the order that never happens
Start with presale questions, the ones shoppers ask before buying. As we explain in shoppers who ask before buying are your best customers, these come from people who are close to paying.
Say your store gets 5 presale questions a day through email and forms, or 150 a month. You reply within a few hours during the week and on Monday for anything from the weekend. Suppose a quarter of those shoppers buy elsewhere before your reply lands. That is 37 lost orders a month.
At an average order of $55, those 37 orders are worth about $2,000 a month, or roughly $24,000 a year.
The second cost: the customer you never keep
A first order is rarely the whole value of a customer. Many shoppers who like a store come back. If a typical customer orders three times over a couple of years, the lost first order is only a third of what walked away.
| Item | Hypothetical figure |
|---|---|
| Lost first orders per month | 37 |
| Average order value | $55 |
| Lost first-order revenue per year | about $24,400 |
| Average repeat orders per customer | 2 more |
| Lost lifetime revenue per year | about $73,000 |
The lifetime figure is the one that surprises most owners. Even if your repeat rate is lower, it is worth including, because the decision to leave happened once and the loss continues for years.
Want a 24/7 AI Employee for Your Website?
See how SpideyChat can answer customer questions, capture leads and handle enquiries while your team is offline.
The third cost: your own time
Now look at the questions that do get answered. If each one takes four minutes to read, check and reply to, 150 presale questions and a similar number of order questions come to about 20 hours a month.
For a founder who also handles buying, photography and marketing, that is half a working week every month spent typing answers that are mostly the same. Sizing questions are the classic example, and we cover them in sizing questions are killing your conversion rate.
Where the slow replies actually happen
Most small stores are not slow all day. They are slow at predictable times:
- Evenings and weekends, when many people shop and nobody is working.
- Busy periods, like a sale or a product launch, when volume jumps.
- Days when the owner is packing orders, at a trade show, or simply off.
This matters because it tells you where to fix things. You do not need to be faster at everything. You need to cover the gaps.
Comparing the ways to reply faster
There are a handful of realistic options, each with a cost:
Reply faster yourself. Free in money, expensive in time and attention, and impossible at 11pm on a Saturday.
Hire help. Even part-time cover is a significant monthly cost for a small store, and it still does not cover every hour.
Canned replies. Tools like the free AI reply generator help you write good answers once and reuse them, which cuts the time per reply.
A website assistant. SpideyChat for ecommerce stores answers presale questions from your product pages and policies at any hour, and captures anything it cannot answer with the shopper's email. You can compare the plans against the figures above.
In the example store, even recovering a fifth of the lost first orders would be worth far more than the cost of an assistant. Your own figures will decide whether that holds for you.
A note on what "fast" means
Speed expectations differ by question. A shopper asking about a delivery problem on an existing order will usually accept a same-day reply. A shopper asking "will this arrive before Saturday?" is deciding right now, and a reply in the evening may be after they have bought elsewhere.
So when you measure response time, split presale questions from order questions. The presale number is the one that costs you sales.
The same logic applies well beyond retail. An estate agency losing buyers to faster agents faces exactly the same arithmetic, as shown in what a missed property enquiry costs an agency.
Running your own figures
You need four numbers: presale questions per month, your best guess at the share lost to slow replies, your average order value, and roughly how many times a good customer orders from you. Multiply through, as in the table above.
If you would rather not build a spreadsheet, the free chatbot ROI calculator will do it. Whatever figure you get, it is usually bigger than the cost of covering the evenings and weekends, which is where most of the loss sits.