How to get repeat buyers

They loved it. They said they'd definitely order again. That was in March, and it's August, and they haven't.

You didn't do anything wrong. You just weren't there on the day they ran out.

Repeat buying is a timing problem before it's a loyalty problem. If you sell something people use up, the second order happens on a date you can calculate, and the whole job is knowing that date and sending one message near it. No points card, no email sequence, no app. A shop doing 30 orders a month that gets a third of its buyers back is doing 40 orders a month by the same effort, and the difference costs you about twenty minutes of admin a week.

Everything below is how to run those twenty minutes.

Repeat buying is a timing problem

Think about what you actually sell. Most small shops sell one of four things, and each has a different clock.

Consumables run out. Soap, coffee, hair oil, spice mix, candles, dog treats. There is a real number of days between purchase and empty, and it's a function of pack size and how often the person uses it. You can work it out with arithmetic.

Wearables wear out or get outgrown. Kids' clothes, socks, work shirts, phone cases. Slower clock, usually seasonal.

Collectibles accumulate. Prints, earrings, plants, books. No clock at all. These people buy again when they see something new, which means your repeat rate depends on how often you release rather than how often they finish.

Services repeat on a schedule the client sets. Haircuts, tutoring, cleaning, photo sessions. The clock is the most predictable of the four and the most commonly ignored.

Almost every piece of advice about "customer retention" is written for the third category and applied to the first, which is why it's useless. If you sell 250ml bottles of hair oil, you do not need a newsletter. You need to know that a 250ml bottle at 5ml a day is fifty days, and to message that person on day forty-three.

Do the arithmetic on your own reorder clock

You can estimate this from the packaging before you ever have data.

Product	Pack	Typical use	Runs out in
Hair oil	250ml	5ml, every other day	~100 days
Body butter	250g	4g daily	~62 days
Ground coffee	250g	18g per cup, 2 cups a day	~7 days
Soy candle	200g	4 hours a week, ~7g/hr	~7 weeks
Spice blend	100g	10g per cook, twice a week	~5 weeks
Dog treats	500g	15g a day	~33 days

Those are starting estimates, not facts about your customers. Halve the confidence and use them to pick a first guess, then replace them with real numbers as soon as you have any.

Getting real numbers is easier than it sounds. Write the send date on the bottom of the jar in pencil, or on the packing slip, and ask people to send you a photo when it's empty. Roughly one in five will, which after nine or ten customers gives you a genuine distribution rather than a guess. That distribution is the single most valuable thing in your business and it costs you a pencil.

The other free source: when someone reorders unprompted, the gap between order one and order two is the clock for that person. Log it. After six of those you'll know your median better than any benchmark could tell you.

Keep the one list almost nobody keeps

Most small sellers can't answer "who bought the lavender one in April?" and that single gap costs more repeat business than anything else.

You need one table. It can be a spreadsheet, a notes file, or the back of a notebook. Six columns:

1. Name
2. How to reach them (the channel they actually reply on, not the one you prefer)
3. What they bought, including the variant
4. Date
5. Amount
6. Anything they said that was specific

Column six is the one people skip and the one that does the work. "Buying it for her mum's arthritis." "Hated the rose one, loved the unscented." "Works nights, delivery before 2pm only." When you message this person in seven weeks, that column is the difference between a message that sounds like you remember them and one that sounds like a mail merge.

If you take orders across DMs, comments and WhatsApp, this list is also the only place they all meet. Platforms lose your history. Accounts get restricted. Your list doesn't, and owning your customer list is worth reading before you find that out the hard way.

The message that gets a second order

Not "how are you finding it?" That's a conversation, and people answer conversations with "good thanks!" and then nothing happens.

The reorder message names the clock, offers the same thing again, and gives them one word to say. It works because it removes every decision except yes or no.

Hi Rina, you got the 250g unscented around the 14th of June. Reckon you're about a week off empty. Want me to put another one aside? Same ₱480, I can send it Thursday.

Four things in there are doing work. The date proves you're not spamming a list. The estimate shows you know the product. "Same as before" removes the variant decision, which is where most reorders stall. And a specific send day makes it concrete.

Send it a week before you think they'll run out, not after. After is too late, because the moment they hit empty is the moment they buy something else in a shop they're already in. A week early lands while they still have some left and no urgency, which sounds worse and converts better, because they're not annoyed and they're not shopping.

Send it once. If they don't reply, don't send a second one two days later. Wait until you genuinely have something new to say, and how to follow up without being annoying covers where that line is.

What Marilou's second orders looked like

Marilou sells homemade chili garlic oil and bagoong from her kitchen in Cebu. A 250ml jar of the chili oil is ₱280, the bagoong is ₱220, and about 70% of her orders are cash on delivery through J&T because that's what her buyers expect. Most of the rest send GCash before she ships.

For her first four months she got almost no repeat orders and assumed the product wasn't good enough. It was. She was invisible between purchases.

What changed was one column in a spreadsheet. She started logging the send date and the variant, then estimated the clock: a 250ml jar of chili oil, used a spoonful most days, is about five weeks. She started messaging at week four.

Her first month of doing that: 34 messages sent, 11 second orders, ₱3,240. Three people replied to say they'd stopped using it, which is useful, and two of those said why. One said the lid leaked in transit, which she fixed with a strip of cling film under the cap and has not heard about since.

The thing she got wrong first: she sent the reorder message to everyone on the list on the same Sunday, including people who'd bought four days earlier. Two of them told her, politely, that it felt like a broadcast. It was. Now she works from the send date column and messages maybe five or six people a week, on different days, which is slower and works far better.

Her repeat orders are also worth more than her first orders. Average first order: ₱280, one jar. Average second order: ₱410, because people who already like it buy two, or add the bagoong.

The second order costs you almost nothing

This is the part worth doing on paper, because it changes what you spend your week on.

Getting Marilou a new customer takes: a post, replies to five people who ask "how much", one who ghosts, one who orders. Call it forty minutes of her time plus whatever the post cost.

Getting Marilou a second order from an existing customer takes: reading one row of a spreadsheet, sending one message. Two minutes. And it converts at roughly one in three rather than one in twenty.

	New customer	Repeat customer
Time per order	~40 min	~2 min
Reply rate	low	high, they know you
Average order value	₱280	₱410
Needs new content	yes	no
Needs trust built	yes	already there

Twenty minutes a week on the second column is worth more than five hours on the first, right up until you run out of past customers. Which is the actual constraint: repeat business scales with how many people have already bought, so it can't be your whole strategy at ten customers. At sixty customers it's most of your income.

That's also why getting your first ten orders has to come first. There's nothing to retain until there is.

Things that quietly kill the second order

None of these are about the product. All of them happen after the money.

Silence between payment and delivery. The buyer paid and then heard nothing for three days. They didn't complain, they just decided you were unreliable. A one-line message when you pack it costs nothing and prevents most of this. "Packed, going out Thursday morning" is the whole thing.

Packaging that arrives looking sad. Not luxury packaging. Intact packaging. A leaked lid, a crushed corner, a label that got wet. People rarely tell you. They just don't come back.

Delivery that cost more than they expected. If the delivery charge appeared at the end and stung, that memory is attached to reordering, not to the first purchase. Buyers do the maths at the second order that they were too excited to do at the first.

Being out of stock the week they wanted it. This one is brutal because it's the moment of highest intent, and you've spent it. If you know a batch is going to be late, message the people whose clock is about to run out, before they message you.

Changing the product without saying. New supplier, new scent, slightly different texture. Say so before they open it. A buyer who was expecting the old one and got the new one silently feels swapped, and that feeling is unrecoverable.

Don't build a loyalty programme

Stamp cards, points, tiers. All of them assume you have enough customers that a percentage-based mechanism is worth the overhead, and all of them require the customer to remember a system.

At your size the mechanism is you remembering them. That's better than any programme and it's the thing a big shop cannot copy.

If you want something structured, the cheapest version that actually works is a standing offer rather than a scheme. "Every third jar, I put in a small one free." No card, no app, no tracking, because you have it in your spreadsheet anyway. Or bundle at the second order: two jars for the price of one and a half, offered only to people who've bought before.

Just do the arithmetic before you offer either. A free jar every third order is a 33% discount on the third order, which on ₱280 is roughly ₱93, and if your margin is ₱120 a jar you've given away most of it. How to run a discount without losing money works through exactly this kind of maths.

The subscription question

Somebody will tell you to offer a subscription. For consumables with a tight clock, coffee especially, it can be right. For most small sellers it's too early, for a reason that has nothing to do with demand.

A subscription is a promise to have stock, every month, forever, on a date you don't control. If you make things in batches in your own kitchen, that promise is the one you'll break, and breaking it costs you the customer you were trying to keep.

A standing arrangement is the version that survives contact with reality. "I'll message you around the 20th of each month, you say yes or not this month." All of the timing benefit, none of the obligation, no payment infrastructure needed. Almost nobody does this and it works better than the thing everybody recommends.

What Sailo won't do for you here

Worth saying plainly, because the gap matters most on exactly this job.

Sailo has no mobile money rail. No GCash, no M-Pesa, no UPI. If half your repeat buyers pay by GCash, the catalogue, the variants, the address and the order all live in the shop, and the payment lives in your phone's notifications where it always did. You confirm it. Nobody confirms it for you. The same is true of bank transfer: Sailo cannot tell you the money landed, only your bank can.

That's fine for a repeat customer, arguably better, because a returning buyer sending you GCash directly is the lowest-friction payment in the world and you've already built the trust that makes it work. It is not fine if you were expecting the reorder to complete itself while you slept.

Card payments do settle themselves, but they carry 1–3% of the goods, and if repeat orders are ₱410 each, that subscription needs a fair number of them before it pays for itself.

Start with the last twenty people who bought

Not a strategy document. A list.

Open whatever record you have of your last twenty orders, even if it's your DMs. Write down the name, the date, and what they bought. Work out the clock for each product using the arithmetic above. Then find the three or four people whose clock has already run past and message them today, individually, naming the thing they bought.

You'll get one or two orders out of that this week from work you did months ago. Then keep the list going forward, because the version of this that compounds is the one where you never again have to reconstruct who bought what.

When the list gets to fifty or sixty names, it turns into something else: a group of people who'll recommend you if you ask properly. Asking customers for referrals is where that goes next.