AI agents for business finance: what they can and can't do
What AI agents can do in bookkeeping, close, collections and compliance today, where a person must stay in charge, and how to judge a product.
Part of the guideOffer pay-later in Nigeria: get paid today, customers pay later
The short answer
Pay-later usually raises both conversion and basket size, but by less than provider marketing suggests. The strongest independent evidence, a randomised study published in 2025, found about a 20% rise in sales, while a Dutch study of hundreds of thousands of orders found pay-later baskets 72% larger than other methods. In Nigeria there is no published data yet, so treat any figure as a hypothesis and measure your own before and after.
Average order value (AOV) is total sales divided by the number of orders. It is the number most pay-later providers lead with, because customers who can spread a cost tend to buy the larger size, the better model or the full course of treatment. This article sorts the evidence into what is reliable and what is sales material, then shows how to estimate the effect for your own business. It sits under our main guide to offering pay-later to Nigerian customers.
What the research actually found
Three kinds of evidence exist, and they deserve different weight.
Controlled experiments
The most reliable study is by Tobias Berg and co-authors, published in the Journal of Financial Economics in 2025. An online retailer offered pay-later to randomly chosen customers, which lets the researchers separate cause from coincidence. Offering it raised sales by about 20%. Most of the gain came from customers with weaker credit who would otherwise have struggled to pay in full, and it was larger for products where the retailer had more pricing power.
Large observational studies
A study by Erasmus University Rotterdam with the Dutch provider in3, published in December 2022, looked at hundreds of thousands of furniture and electronics orders. Customers paying in three instalments spent on average €657.20 per order against €381.20 with other methods, about 72% more. They did not buy more items; they bought more expensive ones. The catch is that customers who choose pay-later may already be planning bigger purchases, so not all of that gap is caused by the payment option.
Provider and retailer case studies
Provider websites often quote basket increases of 30% to over 100%. The Federal Reserve Bank of Richmond accepts that pay-later is linked to higher conversion and average ticket size, but notes that this rests largely on retailer surveys. Such figures come from the best-performing merchants and from firms selling the service. Use them as an upper bound, not a forecast.
Why basket size rises, and when it does not
Basket size goes up when the price, not the desire, is what holds a customer back. That is common for purchases that are necessary or planned but lumpy: a month of medicine, a term's fees, a bridal package, a phone. Nigeria's EFInA 2026 survey found borrowing for coping and consumption rose from 31.7% to 40.8% of formal borrowing between 2023 and 2026, with rent, medical bills and school fees the leading reasons. Those are the same occasions where pay-later at the counter tends to help.
It rises less, or not at all, when:
- tickets are small and frequent, such as a single lunch or a bottle of water;
- most customers can already pay in full and simply switch method;
- the option is hidden, so only customers who were going to buy anyway find it;
- staff push it on everyone, which annoys customers and attracts the wrong kind of sale.
Worked example: does the uplift cover the fee?
Here is a shop with 1,000 orders a month at an AOV of ₦20,000, so ₦20,000,000 in sales at a 25% gross margin. Suppose 150 orders a month move to pay-later at an illustrative 4% merchant fee. The table compares three outcomes for the AOV of those pay-later orders. The fee is charged on the whole pay-later sale, not only on the extra part, which is why a small uplift can lose money.
| Pay-later AOV uplift | Pay-later AOV | Extra sales a month | Extra gross profit (25%) | Fee on pay-later sales (4%) | Net effect |
|---|---|---|---|---|---|
| 10% | ₦22,000 | ₦300,000 | ₦75,000 | ₦132,000 | −₦57,000 |
| 30% | ₦26,000 | ₦900,000 | ₦225,000 | ₦156,000 | +₦69,000 |
| 70% | ₦34,000 | ₦2,100,000 | ₦525,000 | ₦204,000 | +₦321,000 |
This leaves out the other source of gain: customers who would not have bought at all. If even 20 of the 150 pay-later orders are new sales that would otherwise have walked out, at ₦22,000 each, that adds ₦440,000 of sales and ₦110,000 of gross profit, which turns the 10% case positive. That is why the Berg study's result, an increase in total sales and not only in basket size, matters most. For a fuller look at what the fee depends on, see pay-later merchant fees in Nigeria.
How to measure it in your own business
You do not need a research team. You need a clean starting point and a few numbers tracked every month.
- Record three months of baseline figures before you launch: total sales, number of orders, AOV and repeat visits.
- Tag every pay-later sale so it can be reported separately from cash, transfer and card.
- Each month, compare total sales and total orders with the baseline, not just pay-later AOV against cash AOV.
- Count how many pay-later customers are new or had not bought in the last 90 days.
- Subtract the fees and any refunds, then compare the extra gross profit with the fee.
- After 90 days, decide whether to keep, widen or narrow the offer, for example by setting a minimum basket.
Watch for one trap. If pay-later AOV is high but total sales have not moved, your existing big spenders have simply changed how they pay, and you are paying a fee for nothing. The test is total sales and total gross profit, after fees. Inflation also lifts AOV on its own: Nigeria's headline rate was 15.39% in August 2026, so compare against the same months a year earlier with that in mind.
Where Pryme fits
Spaces is Pryme's merchant network for pay-later in Nigeria only: customers pay with a naira spending limit at businesses that have joined it. The credit is lent to the customer by Solid Allianze Microfinance Bank Ltd, a CBN-licensed microfinance bank; Pryme Financial provides the technology. The limit is offered based on how the customer runs their Pryme account, and the customer reviews and accepts the full agreement before any purchase is drawn, so they arrive knowing what they can spend instead of being assessed at your till. Because a Spaces limit can be spent only inside the network, customers holding one have a reason to choose a participating business over one that is not. Whichever provider you use, measure the uplift in your own business with the steps above rather than relying on anyone's published figures.
Whatever bookkeeping you use, the tagging and monthly comparison above are what keep pay-later sales, fees and refunds visible. A Pryme business account comes with a Finance agent that prepares work for a person on your team to approve. The parent guide explains when point-of-sale credit suits a business, and how to tell customers you offer it covers the part that drives most of the uplift: whether customers know the option exists.
Questions people ask
Does buy now pay later increase average order value?
Usually, yes. A Dutch study found pay-later orders were about 72% larger than other orders, though some of that reflects customers who were already planning bigger purchases. A randomised study found total sales rose about 20%.
How much do sales go up when you offer BNPL?
The best independent estimate is about 20%, from a randomised experiment at an online retailer published in 2025. The effect varies by sector, price point and customer base, so measure it in your own business.
Why are provider figures for AOV uplift so high?
They often come from the best-performing merchants and compare pay-later customers with everyone else, without separating customers who would have spent more anyway. Treat them as an upper limit.
How long before I can tell if pay-later is working?
Give it about 90 days. Compare total sales, order count and gross profit after fees with the same period before launch, and with the same months a year earlier to allow for inflation.
Is a higher basket size enough to justify the fee?
Not on its own. The fee applies to the whole pay-later sale, so a small rise in basket size can cost more than it earns. The gain comes mainly from new sales that would not otherwise have happened.
Sources
- Berg, Burg, Keil and Puri: The Economics of Buy Now, Pay Later: A Merchant's Perspective (NBER w33152; Journal of Financial Economics, 2025) · accessed Oct 2026
- in3 / Erasmus University Rotterdam: Paying in 3 instalments leads to 72% higher transaction value (Dec 2022) · accessed Oct 2026
- Federal Reserve Bank of Richmond: Buy Now, Pay Later: Market impact and policy considerations (Economic Brief 25-03) · accessed Oct 2026
- ThisDay: Nigerians turning to survival loans as consumption borrowing jumps (EFInA A2F 2026) · accessed Oct 2026
- Voice of Nigeria: Nigeria's headline inflation eases to 15.39% in August (NBS) · accessed Oct 2026
This article is general information, not financial, legal or tax advice. Products described are provided under the terms on this site and are subject to regulation and eligibility.