A lead-capture checklist for working a trade show booth

What to decide before the doors open, how to capture a lead in ten seconds, and the twenty minutes each night that decide whether any of it converts.

PlaybooksThe Karo team11 min read

Booth teams almost never lose leads at the moment you would expect. They lose them in three specific places: the setup nobody did, the thirty seconds after a good conversation ends, and the week afterwards when nobody can remember which of the two hundred cards was worth a phone call.

None of those is fixed by working harder on the day. They are fixed by deciding a handful of things in advance and then holding to a routine that survives a busy stand. What follows is that routine, in the order you will need it.

Where the leads actually go

It helps to be precise about the leak, because the instinct is to blame the top of the funnel. Teams come back from a show saying footfall was poor. Occasionally that is true. Far more often the conversations happened and the record of them did not survive.

A funnel narrowing from one hundred booth conversations down to seventeen leads that receive a follow-up, with the largest drops occurring at card capture and at follow-up.Conversations100left in a pocket, never photographedCards collected62blurred, cropped, or typed wrongCaptured legibly38no context, so nobody knew what to sayFollowed up17
Illustrative shape, not measured data. The point is where the drops cluster: capture and context, not the number of conversations you have.

Notice that the two largest drops are both about capture rather than about selling. A card that stays in a jacket pocket is indistinguishable, three weeks later, from a conversation that never happened. A card that was photographed badly and typed wrong is worse, because it consumes follow-up effort and produces a bounce.

The final drop is the expensive one. Leads that were captured cleanly but with no record of why they mattered get a generic email, and a generic email to somebody who told you about a specific problem reads as though nobody was listening. We wrote about that particular failure in more detail in why event leads go cold.

Before the doors open

Decide what a lead record has to contain. Not what would be nice to have. What is the minimum that makes a follow-up possible? In practice it is four things: a name, a company, a working email, and one sentence about what the person asked for. If your capture method cannot reliably produce that fourth item, you are collecting contacts rather than leads, and the distinction will become painfully obvious in a fortnight.

Agree who captures, explicitly. On a busy stand, “everyone captures” degrades within an hour into “whoever happens to remember”. Two models work. Either every person captures on their own phone, or one person owns capture and the others physically hand cards over. Both are fine. The failure mode is assuming it will sort itself out.

If you go with everyone capturing, make sure the captures land in one place rather than in five separate phones. Shared team capture is the whole reason our exhibitor setup puts a single balance across multiple seats: the post-show export should be one file, not a merge job.

Set up event tagging before you need it. Whatever tool you use, get the event name attached automatically rather than typed per lead. Typing it per lead means it will be typed three different ways by lunchtime and will be useless for filtering. With Karo that is one message to the bot at the start of the day, after which every card carries the tag. The event capture guide has the specifics.

Test on a genuinely hard card the day before. Not a clean sample card from a stationery shop. Find one of the heavily designed ones with a logo printed across the text, or foil on dark stock, or a QR code where the email should be. You want to know how your setup behaves on a difficult input while there is still time to change plans.

Sort the boring logistics. A power bank per person, because capture dies when phones do. A pocket or an envelope for cards you have already photographed, so you can tell processed from unprocessed. And check that whoever is on the stand actually has the bot linked to their own account, rather than discovering at nine in the morning that one person never finished setup.

At the booth

The constraint nobody plans for is that capture competes directly with the conversation. This is the single most important thing to understand about working a stand. Anything that takes more than roughly ten seconds gets skipped once there are three people waiting, and the leads you skip are disproportionately the good ones, because busy moments are busy for a reason.

A timing comparison. Photographing a card and adding a voice note takes about ten seconds. Typing the same card into a form takes about forty-five seconds and produces no context.PHOTO PLUS VOICE NOTEphoto 3svoice note 7sdone, with contextTYPING IT INTO A FORMreading, tabbing between fields, correcting typos, about 45sand still no record of why the person matteredThe gap matters most when the stand is three deep, which is exactly when the good conversations happen.
Anything that cannot be finished in roughly ten seconds gets skipped once a booth is busy.

So the operating rule is: capture must be finishable without breaking the conversation for long. Everything below follows from that.

  • Photograph, do not transcribe. Typing a card into a form while standing up is the most common way teams fall behind, and once you are behind you never catch up. Take the image and move on.
  • Record the reason immediately. A ten-second voice note the moment they walk away, covering who they were, what they wanted and whether it is real, is worth more than the card itself. Do it before the next conversation starts. Not at lunch, not at the end of the day.
  • Do not qualify while standing up. Resist the urge to sort leads into hot and cold on the floor. You will be wrong more often than you expect, and it costs conversation time you cannot get back. Capture everything, sort in the evening.
  • Keep the physical card if you can. A pocket for scanned cards costs nothing and saves you the one time in fifty that a scan turns out unreadable.
  • Have an answer for the double arrival. When two interesting people turn up at once, take both cards, photograph both, and voice note both in one recording by saying which is which. Trying to run two full conversations in parallel loses both.

One more thing that sounds trivial and is not: say out loud that you will follow up, and when. It sets an expectation you then have to meet, which is the cheapest accountability mechanism available to you.

Each night, not at the end of the show

This is the step that separates teams whose leads convert from teams whose leads become a spreadsheet nobody opens. It takes twenty minutes and it has to happen the same evening.

Three nightly steps totalling twenty minutes: ten minutes reviewing captures, five flagging the most important five leads, and five sending those follow-ups immediately.01Review the day's capturesFix misreads while you still remember the conversation.10 min02Flag the five that matterNot a scoring rubric. The five you would hate to lose.5 min03Send those five nowFirst message in their inbox rather than the fortieth.5 min
Twenty minutes at the hotel beats two hours the following week, because the memories are still separable tonight.

The reason it cannot wait is that memory does not degrade evenly. Around the twentieth conversation of a day, individual details start merging into each other. By the third day of a show you are reconstructing rather than remembering. Tonight you can still tell the operations lead from the procurement manager. Next Tuesday you cannot.

The five-lead limit in the middle step matters too. If you try to prioritise all two hundred you will prioritise none of them. Five is small enough to actually do and large enough to cover the ones that would genuinely hurt to lose.

Send those five their follow-up before the show ends. You will be the first message in their inbox rather than the fortieth, and they will still remember the conversation well enough for a specific reference to land. Everything else can wait for the full export.

After the show

Export once, import once. If several people captured, get everything into one file before it touches the CRM. Deduplicating a spreadsheet is tedious. Merging duplicate records inside a CRM after they have picked up activity history is genuinely painful, and somebody will end up doing it by hand.

Keep the event tag through the import. Six months from now, the only way to answer “was that show worth going to?” is if every lead from it is still identifiable as a group. Most teams discover they cannot answer this the first time a finance director asks. Map the tag to a campaign or a source field, whichever your CRM uses, and check it survived. Our HubSpot import guide walks through the column mapping if you want a worked example.

Write down what broke. Two lines in a shared document while it is still fresh. The next show is the same show with different carpet, and the same three things will break again unless somebody recorded them. This is the highest-return five minutes of the entire trip and almost nobody does it.

Decide what the show cost per usable lead. Not per card collected. Per lead that a salesperson would actually accept. That number is the only honest input into whether you book the stand again next year, and it is only available to you if you kept the event tag.

What the organiser’s lead file will not tell you

Most shows sell some form of lead retrieval, and most exhibitors treat the file it produces as the record of the event. It is worth being clear about what that file actually is, because assuming it covers you is how teams end up with nothing.

It contains whatever each attendee typed into a registration form, possibly months before the show. Job titles in it are frequently one role out of date. Email addresses are often the generic company one rather than the person’s direct address. And if your staff scanned badges indiscriminately to hit a number, it contains a large quantity of people who walked past your stand without stopping, indistinguishable from the people who had a twenty minute conversation.

What it never contains is anything either party said. There is no field for the fact that somebody is replacing their system in the first quarter, or that the person you spoke to does not hold the budget.

Use it, because it is usually already paid for. But reconcile it against your own captures rather than the other way around: your list tells you who mattered, theirs helps fill in a missing email address. If it arrives two weeks after the show, which is common, your own follow-ups should already have gone out. The trade show capture guide goes into the practical differences.

Adjusting for team size

If you are on your own, your bottleneck is that you cannot talk and capture simultaneously. Lean harder on the voice note and accept that some cards will get photographed in a batch during a quiet ten minutes. Batch the photos if you must, but never batch the voice notes. The photo is recoverable later; the context is not.

If you are a team of ten, your bottleneck is consistency. Ten people will produce ten standards unless you agree one beforehand, and the person who captures least carefully sets the quality of the merged list. Spend fifteen minutes on the morning of day one walking through exactly what a good capture looks like, including a demonstration.

If you are somewhere in between, which most booths are, appoint one person to own the nightly review even if everyone captures. Shared ownership of the evening routine means it does not happen.

The short version

Decide the minimum lead record and who captures it. Set up event tagging before the doors open and test on a difficult card. At the booth, photograph rather than type, and record why the person mattered in the ten seconds after they walk away. Review the same night, flag five, and send those five before the show ends. Export once, keep the event tag through the import, and write down what broke.

If you want the capture half handled inside the chat app you already have open, Karo starts with 20 free scans, which is enough to work one real day and find out whether it holds up on your cards.

Frequently asked questions

How many people do I need on a booth to capture leads properly?+

Capture is not really a headcount problem, it is an ownership problem. One person can capture well on their own if they lean on voice notes. A team of ten can capture badly if nobody agreed a standard. What matters is that someone explicitly owns capture, and someone explicitly owns the nightly review, even on a small stand.

Should I use the organiser's badge scanner instead?+

Use it if it is included, but do not rely on it alone. Badge scanners give you whatever the attendee put into the registration form, which is often a job title from two roles ago and a generic company address. They also give you no record of your conversation. The two work fine together: scan the badge for the data, voice note the context.

What do I do about cards I cannot read at all?+

Keep the physical card and flag the capture for review rather than guessing. Guessing produces a bounced email and a lead that looks followed up when it was not. Karo keeps the original photo and audio even when extraction fails, so a bad scan stays recoverable.

Is it rude to photograph someone's card in front of them?+

In our experience it reads as the opposite. Photographing a card in front of somebody signals that you intend to do something with it, whereas putting it in a pocket signals the opposite. Saying what you are doing helps: something like, I am saving this now so it does not get lost.

How soon should the first follow-up go out?+

For the handful of leads that genuinely matter, before the show ends. For everyone else, within a few days of getting home. The specific number of hours matters less than being early enough that a specific reference to your conversation still means something to them.

What if my team refuses to write notes?+

That is usually a sign that the method is too slow rather than that the team is careless. Typing notes competes with talking and loses. Speaking a ten-second voice note does not, because it takes about as long as saying goodbye and requires no decisions about which field the information belongs in.

Do I need the physical cards after scanning them?+

Not for the data, but keeping them until the export is verified costs nothing and covers the rare unreadable scan. Once the leads are in your CRM and you have checked a sample, the paper has done its job.

Keep reading