When I was running my restaurant, I spent almost all my energy on one question: how do I get more people in the door?

Not a bad question. But it's the wrong one to spend all your time on, because getting someone in the door once is the most expensive customer you will ever buy. The flyer, the event, the ad, the free sample — all of that money goes into a first visit. And then most of those people walk out and you never see them again, and you go run another promotion to replace them.

That's not a marketing strategy. That's a treadmill.

The restaurants that actually make it aren't the ones who are best at getting strangers in. They're the ones who are best at what happens after.

The funnel, in plain English

Marketing people love the word funnel. It sounds complicated. It isn't. For a restaurant it's four steps:

1. Stranger — they've never heard of you

2. First-timer — they came once

3. Repeat — they came back

4. Regular — you know their order

And then there's a fifth one that most people forget:

5. Advocate — they bring other people with them

Every dollar you spend on marketing is trying to move somebody from one step to the next. The mistake is spending 90% of it on step one.

The second visit is the whole ballgame

Here's what I learned the hard way: the gap between visit one and visit two is the widest gap in the whole funnel.

Somebody who comes twice is a fundamentally different person than somebody who came once. Coming once can be an accident — they were walking by, their friend picked the place, they had a coupon. Coming twice is a decision. They chose you. Once somebody has chosen you on purpose, getting them to visit three, four, and ten is dramatically easier.

So if you only fix one thing about your marketing, fix the 48 hours after a first visit. That's the highest-leverage window you have and almost nobody uses it.

What that looked like for us:

• Get something in their hand on the way out. A card in the bag with a reason to come back. Not "10% off," which is forgettable — something specific and slightly weird, like a free side on your next order. Specific beats generous.

• Get a way to reach them. A phone number for order-ready texts, an email for the online order receipt, a loyalty signup at the register. If the only place a customer exists is in your memory, you don't have a customer. You have a nice afternoon.

• Give them a low-friction second option. A lot of people who liked you won't come back in for a while, but they'll order online next Thursday when they don't feel like cooking. Make sure they know you do that.

None of that costs real money. It's all just remembering that the person in front of you is worth more than the next stranger you're trying to buy.

Stop trying to turn everybody into a weekly regular

This is the part I got wrong for a long time.

I used to think the goal was to get everybody coming in every week. So we'd blast everyone with the same promotions at the same frequency, and we'd wonder why people tuned us out.

The truth is your customers already know what they are. Some people are Tuesday-lunch people. Some are once-a-month date night people. Some are twice-a-year, my-mom's-in-town people. Those are three completely different customers and none of them are wrong.

Your job isn't to force the birthday person into a weekly cadence. Your job is to be the place in whatever category they've already put you in — and then talk to them at the speed they actually move.

So cast a wide net at the top, and then let people self-sort. Watch what they actually do, not what you wish they'd do. The weekly guy gets a different message than the twice-a-year guy. The twice-a-year guy doesn't need a text every Friday; he needs one really good text in November before the holidays.

The moment I stopped trying to convert everybody into the same customer, my marketing got a lot cheaper and a lot more effective.

What to actually do at each stage

Stranger → First visit. This is the part everyone already focuses on, so I'll keep it short. Be findable (your Google profile matters more than your website), be reviewed, and give people a reason that's happening this week. Events are the cheapest version of this — I wrote about the grand-opening-every-month approach here, and about why Google reviews do more of this work than most owners realize here.

First visit → Second visit. The 48-hour window above. Reason to return, a way to reach them, an easy online option.

Second → Regular. This one is mostly about recognition. People come back to places where they feel known. Learn names. Learn orders. If you can't do that from memory at your volume, your POS should be doing it for you — the person at the counter should be able to see "this is their 6th order, they always get the same thing" and say something about it. A loyalty program helps here too, but the points are honestly the smaller half of it. The bigger half is that a loyalty program forces you to actually track who's who.

Regular → Advocate. Just ask. Regulars want you to succeed and most of them have never been asked to do anything about it. Ask for the review. Ask them to bring somebody. Give them a reason to — a free item for a friend's first visit costs you a few bucks and buys you a first-timer who arrives already trusting you, which is the best kind.

You can't work a funnel you can't see

Everything above depends on one thing: knowing who's new and who's not.

That's where most independent restaurants are flying blind. You've got a register that tells you what you sold today and nothing about who bought it. So every customer looks identical, which means you end up treating them identically, which is exactly the problem.

This is one of the things we built Avocado around. Your POS already touches every transaction — it should be telling you what percentage of today's tickets were repeat customers, who hasn't been in for 60 days, and which of your regulars just went quiet. It runs the loyalty program, sends the order-ready and come-back texts, and keeps the customer list that makes all of this possible instead of theoretical.

You don't need a marketing department. You need to stop throwing away the data walking through your door every day.

If you do one thing this week

Pick the second-visit window. Just that one.

Figure out what a first-timer leaves with, and what happens in the two days after. If the answer is "nothing," you've found the cheapest improvement available to you — and it costs less than the ad you were about to run.