Guides
How to tell if more brewery and taproom customers come back
Count first visits and repeat visits, compare the same weekdays, and read the result so you know whether more taproom customers really come back.
What to take away
- A taproom can see more people and still win no more customers. What counts is how many first visits come back.
- Count first visits and repeat visits separately, on one weekday and one day part.
- Compare like with likethe same weekday, in the same quarter, before and after the change.
- If the prior quarter was flat, treat the change as weather until you see it twice.
- A repeat visit needs a name to match on, so choose the matching method before the count starts.
- Reviews explain a flat result, provided you ask for them within the rules.
Why first visits that return are the number to count
Count first visits that return, not total traffic. A busy week can come from the weather, an event or one large party, and none of those tells you whether strangers turned into customers. The return rate of first visits is the count a change in service can actually move.
Reviews are how most first visits are decided for people who have never been in, so the count and the reviews belong together. The count says whether people come back. The reviews say why they did or did not, in the guest's own words.
The count in four steps
Compare like with like: the same weekday, in the same quarter, before and after the change. Everything else in the method follows from that rule, and it keeps the comparison fair. A promotion judged against the week it ran will almost always look successful.
Measuring conversion honestly
- Compare like with like, same weekday
- Note festivals, holidays, closures, weather
- Watch party size, not just headcount
- Look for a trend, not a month
- Ask for a mechanism in one sentence
- Pick one weekday and one day part, say Thursday from 6pm to 9pm.
- Count covers and count first visits. Ask at payment: "Have you been in before?" A booking note or a till prompt does the same job.
- Count repeat visits, which needs a name to match on. Loyalty scans or a card on file are the clean way. Without either, a paper tally works for four weeks.
- Compare the four Thursdays after the change with the four before it, and with the same four Thursdays a year earlier.
Step three needs a name to match on, because a repeat visit cannot be counted from a till total alone. Pick the cheapest way to match names that a bartender can run in a few seconds during a rush, and write down which one you chose.
A worked example
Suppose 200 people make a first visit in April and 30 of them have come back by July. That is 15 percent. The same count a year earlier gives you the comparison that matters, because it holds the season steady.
| Period | First visits | Came back | Return rate |
|---|---|---|---|
| This April's first visits | 200 | 30 | 15 percent |
| Last April's first visits | 200 | 22 | 11 percent |
If the same count last April was 22 of 200, you moved from 11 percent to 15 percent. If the prior quarter was flat, treat the change as weather until you see it twice. Write the result down as counts, not as a feeling, and file it with the date.
Four mistakes that spoil the count
Four mistakes spoil a return count, and each has a fix that costs nothing. All four come from skipping one of the rules above, so check the list before you trust a number.
| Mistake | What it does | Fix |
|---|---|---|
| Judging a promotion against the week it ran | It will almost always look successful | Compare four Thursdays before and after, and last year's four |
| Counting total traffic | Weather, an event or one large party lifts it | Count first visits that return |
| Mixing weekdays and day parts | Compares unlike nights | Fix one weekday and one day part |
| Matching on the till total | A repeat visit cannot be counted from it | Loyalty scan, card on file or a paper tally |
Reading the result without fooling yourself
A change is real when it shows up more than once, on the same weekday, against the same quarter a year earlier. One good Thursday is weather. Four good Thursdays against the four before them, and against last year's four, is a result worth acting on.
Then match the shape of the week to the job. Winning more customers is only useful if you can serve them, and a rate that rises on nights already at the limit of room and crew is not good news yet.
Check the constraint first
Are busy nights already maxed out?
More first visits land on your worst experience
Is midweek quiet and weekends full?
| What your week looks like | The job |
|---|---|
| Busy nights already max out room and crew | Fix capacity first, because new first visits meet your worst experience |
| Midweek quiet, weekends full | Move demand rather than create it |
| Both quiet | Creating demand is the job |
| Consistently full across the week | Not a customer question any more: see the routes to more volume |
The last row is a structural decision with its own approvals and payroll, so it belongs outside a customer count. A guest who comes back is counted here. A guest who brings friends is a different measure, covered in turning bar moments into referrals.
When the rate is flat, walk in as a stranger
A flat return rate points to something a guest met, so go and meet it. Walk in as someone who has never been here and work through five steps in order. Do it twice a year, on a busy night rather than a quiet one, because problems appear under load.
The twice-yearly audit
- Send a first-timer in alone
- Record all five moments in writing
- Listen without explaining
- Fix only the worst moment
- Repeat in six months
- Find the hours.Check the website, the map listing and the social profile before you leave home. Write down any place they disagree.
- Stand inside the door for thirty seconds.Time how long before someone acknowledges you, and note whether you can tell where to order.
- Ask what to drink.Say what you usually enjoy. Listen for a question back, then for a suggestion you could repeat to a friend.
- Raise one small problem.A warm glass, a beer that is not what you expected. Watch how it is handled, and how long it takes.
- Ask what would bring you back.If nothing comes to mind, there is no reason to return yet.
Step three is the hard part and the whole value. The instinct to explain is strong and it is exactly what a first time guest never gets.
The five moments
- Finding you and knowing you are open
- First thirty seconds inside
- The first recommendation
- The correction
- The reason to come back
The walk covers five guest moments, from finding you to the reason to come back. Four happen inside the building and involve staff, so a low rate is a training subject as much as a marketing one. Hiring and training the crew is where it gets fixed.
Fixed by the building, or fixable this month
Some of what a stranger notices is decided by the building, not the crew. Where a queue forms and whether it blocks the entrance are set by the fit out. Separate the two lists honestly, so a layout problem is not blamed on service.
Fixed until you spend money on the building:
What the room does
- Where the entrance sits in relation to the bar.
- Whether the bar can see the door.
- The walk from bar to far table.
Fixable this month with staff, signage or furniture:
What the room does
- A queue that blocks the entrance. Move it with a sign or a second till.
- A blind spot between bar and door. Put a bartender on the floor side at peak.
- Hours that are wrong online. Fix them today.
Fix what is fixable and staff deliberately around the rest. A room the bar cannot see needs someone on the floor, and no amount of training substitutes for that.
Price is part of the same first impression. Guests compare prices within your list far more than against other venues, and they read value from the glass, the fill level and whether the pour is consistent. Check whether a perception problem is really a consistency problem before reaching for a discount.
What reviews add to the count
Reviews explain a flat count but do not replace it. Group the quarter's reviews by the guest moment they mention, then fix the one that repeats. Reply to every review within a week and do not argue.
Review solicitation rules
- Ask everyone in the same way
- Never only the people who seem happy
- Disclose clearly if anything is offered
- Check state alcohol board restrictions
Asking for reviews has rules. The FTC's guide to soliciting reviews tells businesses not to ask only customers they think will leave positive ones. In practice: ask everyone in the same way, never only the people who seem happy, and disclose clearly if anything is offered in exchange.
Incentives are treated seriously. Under the Consumer Reviews and Testimonials Rule, a review that a business pays for or rewards with an incentive is also a consumer testimonial. Your state alcohol beverage control board may also restrict what an alcohol licensee can offer as an inducement, so ask before designing anything with a reward attached.
Common questions
Should I advertise to reach new customers?
Only after the guest moments are working, because advertising into a room that loses first time guests is paying to lose them faster. Fix the conversion, then raise the traffic.
What is the fastest thing I can fix?
Hours that are wrong somewhere. It costs nothing, it takes an afternoon, and it silently turns people away who had already decided to come.
How long before I trust a change in the return rate?
Treat the change as weather until you see it twice. Compare the same weekday and day part against the quarter before and against the same quarter a year earlier.
Do events bring lasting customers?
Sometimes, and the way to tell is to look weeks later rather than on the night. Events that work bring first time visitors who return, so the measure that matters shows up in the following month.
Is a loyalty scheme worth it?
Only if a bartender can run it in a few seconds during a rush, and only if your state permits the mechanic. Otherwise it becomes a thing that is done inconsistently and then quietly dropped.
