If you installed Riffl this week, scanned a couple of receipts and found a tidy list of transactions, you have seen roughly a fifth of it. That is my fault, not yours — nothing in the app currently tells you which habits switch the rest on. This is that missing page.
Scan the receipt even when the bank already told you.
This is the one that matters most, so it goes first. A bank alert gives Riffl one number: the amount. A receipt gives it the items. Those are not the same thing, and almost everything interesting in the app runs on items — the basket X-ray, your personal inflation rate, the impulse signals, the price-creep findings. Scan only alerts and you get an accurate ledger of what left your account. Scan the receipts too and you get the thing that tells you something you did not already know. When you buy something and the alert lands before you have left the counter, take the paper anyway and snap it. The alert and the receipt merge into one expense — Riffl detects the duplicate rather than counting it twice.
Set a budget now, not on the 15th.
A budget is not just a cap — it is what the pace projection measures against. Riffl takes how far through the month you are, how much has gone, and projects where you land, then tells you whether that is under or over and roughly which day a category crosses its cap. All of which needs a budget to exist, and needs it to have been there while the spending happened. Set one halfway through and the projection has only seen half your month, so it is working from half the evidence. Ten minutes today is worth more than an hour later. It also counts the days you have held it — the cap is yours, not one Riffl picked for you.
Then a savings goal, because the two are connected.
Set a target and Riffl works backwards from it — whether you are on pace or falling short. The part worth knowing is what it does with the budget: if you are projected to come in under this month, it offers to set that under-budget money aside toward the goal, so an ordinary good month turns into something you can see. To be explicit, because it matters — Riffl records the transfer, it does not make it. It never touches, moves or holds your money, and it never has access to an account that could.
The impulse engine needs five trips to the same shop.
Riffl works out what you actually go to a shop for by watching what you buy there repeatedly. An item you buy on 60% or more of your trips is a staple. One you buy on under 30% is an extra — the impulse candidate. But it refuses to guess: until there are five separate trips to that merchant, it has no baseline and it will say so rather than invent one. Five trips to one shop is worth more here than one trip to five shops.
Then it can tell you when you overspend, not just that you do.
Once four or more extras have accumulated, Riffl builds a time signature — what share of your impulse buys happen in the evening, which part of the day is your peak, and whether they cluster in the days right after money comes in. "You spend too much on snacks" is not useful. "Four in five of your unplanned buys happen after 7pm, mostly in the three days after money lands" is a thing you can actually do something about.
Ajo, and the money you lent Tunde.
Riffl keeps a counterparty ledger — who owes you, who you owe, and what has been repaid so far. It is genuinely useful and it starts completely empty, because nothing can populate it until you log the first one. Contributed to an ajo this month, lent a colleague ₦20,000, still owed for the shared bolt fare? Put one in. Log a repayment against it later and Riffl matches it to the original and works out what is left. Nobody discovers this feature by accident, which is why almost none of you have used it.
Tag one expense #business and a tax pack appears.
If you run anything on the side, tag those expenses #business. Riffl keeps a running business-only summary, and the Reports tab will build a tax pack over any range you pick — a quarter, a half, a full tax year, or dates you choose — as a PDF statement plus a CSV your accountant can actually open. There is a business-only switch on it, so the pack contains your business spending and nothing about your groceries. It also backs out the VAT: Nigerian prices are quoted VAT-inclusive, so Riffl takes the 7.5% portion out of each amount and totals it for you, marking the exempt categories rather than pretending they carry it. The whole thing is generated on the phone — nothing is uploaded to be rendered. The tag is the entire setup; there is no separate business mode to turn on.
Correct a category once and it sticks to that merchant.
When Riffl files something wrong, changing it is not just a one-off edit. It records the correction against that merchant, and every future transaction from them is categorised your way. So the fifteen seconds you spend fixing "Total Filling Station" from Shopping to Fuel is fifteen seconds spent once, not every month. This is the cheapest possible investment in the app and it is invisible — nothing tells you the correction was learned, it just quietly stops being wrong.
Cash is the money that vanishes.
No alert, and often no paper either. Cash spending is the single biggest hole in most people’s picture of their own money, and it is the one place Riffl cannot help unless you help it first. If you paid cash and got a receipt, scan it. If there was no receipt at all, add it by hand — it takes about ten seconds and it is the difference between a monthly total that is roughly right and one that is quietly missing a third of your life.
Turn backup on before you need it.
Your data lives encrypted on your phone. That is the privacy promise and it is a real one — but the flip side is that a lost, stolen or factory-reset phone loses everything with it, and there is no copy on my side to restore from, because by design I cannot read it. Settings has a backup section. Turning it on takes under a minute. Do it now rather than at the point where it would have mattered.
What unlocks when.
Roughly, so you know what you are waiting for rather than concluding it is missing. Your weekly review appears on Sunday — weeks run Monday to Sunday, and it needs some spending in that week to review. The basket breakdown for a shop needs five trips to that shop. Impulse time signatures need four extras. Price history for an item needs three purchases of it. Your personal inflation rate is the slowest: it compares median prices across two different calendar months, so if you started in August, it has something to say to you in September and not before. That one cannot be rushed, and it is the one most worth waiting for.
None of this is a checklist you have to complete. But the first three — scan receipts and not just alerts, set a budget, log one ajo or loan — are what separate an app that shows you a list from an app that tells you something. Everything else follows from having items in it.
If something here does not match what you are seeing, that is a report and I want it. Settings → Report a bug or idea opens WhatsApp with the questions already written, or just reply to the email that sent you here.
— Olamide