Your first sale is the expensive one. Most product businesses never set up the second.

Most marketing advice for product businesses is about getting the first sale. Ads, photos, pricing, the shop, the launch, the market stall, the shelf.

Almost none of it is about the second one, which is where the money actually is.

The sum nobody runs

Say it costs you thirty dollars in advertising, discounting and your own time to get someone to buy a forty dollar product for the first time. On that sale you made very little, and possibly nothing.

If that person buys three more times over the next year, that customer was excellent. If they buy once and vanish, you ran a charity.

The uncomfortable part is that both outcomes look identical on the day of the sale. You cannot tell them apart from your till, your month, or your revenue total. The difference only shows up later, and only if you set something up in advance to notice it.

You probably do not know who bought from you

This is the actual bottleneck, and it is far more common than owners expect.

If you sell in person, at a market, at a fair, in a shop, through a card machine, then you have money and no customer. The transaction told you nothing you can act on. The person who has bought your best selling item four times is, as far as your records go, a stranger.

Selling online is usually better, because an email address arrives attached to the order. But plenty of product businesses have that data sitting inside a platform they never export from, never sort, and never actually send anything to.

Either way the problem is the same. You cannot sell to someone a second time if you have no way to reach them.

The three things worth capturing

You do not need a loyalty programme or a database project. You need three things.

• A way to reach them. Email is still the best of a mediocre set of options, because you own the list. A phone number is better still if they will give it and you will use it properly.
• What they bought. Not a full purchase history. Just enough to know which thing brought them in, because that is the thing you reference next time.
• When they bought it. Almost everything you might send is really a question about timing. A candle, a bag of coffee and a skincare product all run out on a schedule, and the schedule is the whole opportunity.

A spreadsheet does all three. It is not elegant and it works.

What to actually send

This is where most attempts die. The owner sets up a list, has nothing to say, and sends a discount. Then another discount. Then they have trained everybody to wait for discounts.

Better options, roughly in order of how well they work:

• A reorder nudge timed to when the thing runs out. Not a sale. A reminder at the right moment. This is the single highest value message a product business can send.
• The thing that goes with the thing they bought. Obvious, underused, and it does not require a price cut.
• New arrivals, sent to people who bought that category. Not to everyone. To the forty people it is actually relevant to.
• A restock alert for the item they wanted and could not have. These do unusually well, because the intent already existed and you are not having to create it.

None of these require you to be clever or funny, or to have a content strategy. They require you to know what someone bought and when.

If you sell through a shelf you do not own

Wholesale and stockists complicate this, because the shop owns the customer relationship and you only get the order.

You still have two routes. The packaging is one. It goes into the customer's home and stays there, and it is the only direct line you have, so it is worth carrying something more useful than a logo and a web address.

The second is to treat the stockists themselves as the customers worth keeping, because they are. A shop that reorders every quarter for three years is worth more than almost any individual shopper, and hardly anyone markets to their stockists at all.

Where to start

Pick the narrowest version of this that you can finish this week.

Start collecting email addresses at the point of sale, with a reason attached that is not a discount. Then write down, for your single best selling product, roughly how long it lasts. Then send one message to the people who bought it that long ago.

That is the whole project. One product, one message, one list. If it works, the version with the proper tools and the sorting and the automation is worth building. If it does not, you have lost a week rather than a quarter.