When you don't need a paid search engine

The front page of this blog says you'll find honest guides here, "including when Findalo isn't the right option". That went unfulfilled for a long time. This is the post that fulfils it.

Here are six situations in which you shouldn't buy a paid search engine — not ours and not anyone's — along with what's worth doing instead in each. If you recognise yourself in any of them, you've just saved between €180 and €1,800 a year.

1. You have fewer than two hundred products and your customers know them

With a small, stable catalogue, people don't search: they browse. They know you have four categories and they go straight there. Search gets used rarely and, when it does, it almost always finds something, because there isn't much to get lost in.

Do this instead: fix the navigation. Categories with names that mean something, a menu that fits on a phone screen, and good photos. It'll pay off far more.

The exception: if those two hundred products are technical and get ordered by code — spares, components, consumables — then people do search, and a lot. There, the number of SKUs isn't what decides.

2. Your traffic is low

The sum is uncomfortable and it has to be done. A paid search engine is justified if the money it moves exceeds what it costs. That money depends on how many people search, how much they buy, and how much each purchase is worth.

With few visits a month, the number of people who get as far as using search is so small that no improvement, however large, shows up in the till. And even if it did, you couldn't tell it apart from the normal noise of a quiet month.

Do this instead: traffic. Whatever brings people in. An excellent search engine in an empty store is still an empty store.

3. Your platform's search hasn't run out of road yet

Platforms have improved their native search, and most stores have never touched its settings. There's a fair amount of free headroom before you pay for anything.

The settings that pay off most and that almost nobody looks at: the minimum word length that gets searched, the weight of each field — if reference is at zero and your customers search by reference, there's your problem — the ignored-words list, and product visibility, because one marked "catalogue only" will never appear.

Do this instead: run the fifteen checks. If you fail fewer than four, your search is fine and you don't need another one.

4. Your product pages are empty

This is the most important of the six and the one that prevents the most wasted money.

If your products have a name, a price and two lines of description, there is no search engine in the world that will find one by what it does. The information doesn't exist. It isn't badly indexed: it isn't there.

You'll notice it when your customers search by use, by problem or by compatibility — "for sensitive skin", "compatible with", "for outdoor use" — and nothing comes up, even though you stock it.

Do this instead: write up your fifty best sellers. What they're for, who they're for, what they're compatible with, what size they are. It's boring work and it's the best paid: it improves the search you already have, it improves Google, and it improves what an AI assistant can say about you.

5. Your problem is data, not search

If the stock your site shows isn't the real one, if prices arrive late from your management system, or if the same product is listed three times under different names, a new search engine will find those problems faster and with more people watching.

That's useful, but it isn't what you were buying.

Do this instead: fix the sync with your management system, merge the duplicates, then come back. We read whatever is there; if what's there is a mess, it shows.

6. You sell few things, or you sell subscriptions

A store with one product, one subscription or one service doesn't have a search problem. It has a persuasion problem, and that gets solved on the page, not in a text box.

Do this instead: everything else.

So when does it make sense?

So this isn't only a list of noes, here's the other side. A paid search engine starts to pay when more than one of these is true:

• You're past a few thousand SKUs and still growing.
• You sell in several languages, which is where native search breaks first.
• Your customers search by code, with hyphens, leading zeros and variant suffixes.
• You have the list of what gets searched and not found, and it's long and full of things you do stock.
• You sell to businesses and each customer has to see their own price and their own catalogue.
• You have enough traffic for a 10% improvement in search to show up in the till.

With just one, probably not. With three, almost certainly yes.

Why we write this when we sell search engines

Because customers who leave after three months cost more than they bring in, and because a significant part of our sales work consists of saying no to people we could have charged.

And because the alternative — sell to everyone and let the product fend for itself — works once per customer. This works every time.

If you've read this far and you think you're in one of the six cases, don't write to us: do what your case says. If you think you're not, write to us with your store's URL and we'll tell you what we see, including if we see that you don't need us.