Trade Show ROI: How B2B Teams Actually Measure Event Pipeline (2026)

Trade show ROI is the number that decides next year's event budget. Most teams cannot calculate it.

They know the booth cost six figures. They know roughly how many badges got scanned. Then the trail goes cold. The leads sat in a CSV for two days, landed in the CRM without a campaign source, and got worked by reps who did not tag the event. Six months later the CMO asks "what did Q1 events produce" and the honest answer is a shrug dressed up as a dashboard.

This is a measurement problem, not a channel problem. The channel works. The plumbing under it leaks.

I run Tendro, so factor in the bias. This piece is the operating model for measuring event ROI: what good looks like, which attribution to pick, the real benchmarks, what a lead costs, and why so many teams underreport. The Tendro pitch is one section at the end, flagged.

What counts as good ROI for a B2B trade show?

Good trade show ROI means closed-won revenue that beats all-in event spend. Many teams set an internal pipeline-to-cost target near 3-to-1.

That target is a working convention, not a law, and plenty of teams aim higher or lower. The honest version depends on your sales cycle, your deal size, and whether you measure sourced or influenced pipeline. A six-month enterprise cycle will not show closed-won 30 days after a show. So most teams report ROI in two passes: a pipeline number at 30 days, a closed-won number at 90 to 180 days.

The mistake is comparing the wrong things. Teams stack the full booth cost against first-touch closed-won only, ignore every deal the event influenced later, and conclude events do not pay. The reverse mistake is crediting the event for every deal anyone at the show ever touched, which inflates ROI into fiction. Good measurement sits between those, and getting there requires picking your attribution rules before the show, not arguing about them after.

One reason the conversation is muddy: events eat a large share of B2B marketing budgets, so the stakes of getting the number wrong are high. CEIR's 2018 Marketing Spend Decision Report found that, on average, companies that exhibit allocate around 41% of their annual marketing budget to exhibitions. That figure is a few years old, so treat it as directional rather than current. The point holds regardless of the exact percent: this is the single biggest line item most field marketing teams defend, and "trust me, the booth was busy" does not survive a budget review.

How do you actually measure trade show ROI?

Track all-in spend, captured leads, MQLs, SQLs, sourced pipeline, influenced pipeline, and closed-won. Pull the same seven numbers every show.

The model is boring on purpose. Boring is repeatable, and repeatable is what lets you compare this show to last show. Here are the seven inputs and what each one actually requires.

1. All-in spend. Booth build, shipping, drayage, staff travel, hotels, swag, the lead-capture tool, the follow-up email tool, the post-event client dinner. The number you defend to finance. Most teams undercount this by leaving out travel and the soft costs, which flatters ROI and gets caught the first time the CFO asks for receipts.

2. Captured leads. Total records off the floor before dedup. Badge scans, business cards, QR claims, handwritten notes. This is your raw top of funnel.

3. MQLs. Post-dedup, post-ICP-filter. The leads marketing will stand behind as real.

4. SQLs. The subset sales accepts and agrees to work. The gap between MQL and SQL is where Marketing-Sales disagreements live.

5. Sourced pipeline. Opportunities where the event is the first-touch origin, tagged via Primary Campaign Source in Salesforce or the equivalent in HubSpot, Marketo, or Pardot. The cleanest, most finance-respected number.

6. Influenced pipeline. Opportunities the event touched at any point in the deal but did not originate. A multi-touch model credits the event a slice. The halo effect lives here.

7. Closed-won. Trace the won deals back through their campaign sources at 30, 60, 90, and 180 days. The event either shows up in the path or it does not.

The reporting cadence that works: pipeline at 30 days, revisited closed-won at 90 and 180. Run the same template for every show. The first time you do this it takes a day. After that it is a saved report.

Where the model breaks in practice, every time:

• The spend number is incomplete because nobody logged the dinner or the travel.
• The captured-lead count is overstated because the organizer kiosk double-counts and the export has dupes.
• The MQL count is opinionated because Marketing and Sales never agreed on the bar.
• The closed-won attribution is contested because sales credits the SDR and marketing credits the event.

None of those are tooling failures. They are discipline failures. The template gives you the shape. Clean source data and agreed definitions do the rest, which is exactly where most of the leakage hides.

What is the difference between sourced and influenced pipeline?

Sourced pipeline is deals where the event was the first touch. Influenced pipeline is deals the event touched at any point but did not originate.

This distinction is the whole game in event attribution, and it is where most reporting arguments start. Sourced is a single-touch idea: one campaign gets full credit as the origin. In Salesforce that is the Primary Campaign Source field on the opportunity. It is binary and clean, which is why finance likes it. It also undercounts events badly, because a trade show is rarely the literal first time a buyer heard of you. They saw an ad, read a blog, then stopped by the booth. Pure sourced attribution gives the booth zero credit for that deal.

Influenced pipeline fixes the undercount. It counts every deal the event touched anywhere in the journey, then a multi-touch model decides how much credit the touch gets. The risk flips the other way: influenced numbers are generous, and if you report influenced pipeline as if it were sourced, you are double-counting across every channel that touched the same deals. Sum every channel's influenced number and you will "produce" several times your actual pipeline.

The operator answer is to report both, label them clearly, and never blend them in one figure. Sourced is your floor: the pipeline you can defend as event-originated. Influenced is your ceiling: the pipeline the event helped move. The truth for budget decisions sits between the two, and showing both is more credible than picking the one that flatters the show.

A practical note on the booth conversation. A scanned-but-uncontextualized badge is hard to attribute to anything later, because nobody knows whether it was a real conversation or someone grabbing swag. The qualification you capture at the booth (hot, warm, cold, the ask, the next step) is what lets you separate sourced from influenced honestly three months later. Attribution is a data problem before it is a modeling problem. More on the capture side in the event leads to CRM hub.

Which attribution model should you use for events?

Pair one sourced model (Primary Campaign Source) with one multi-touch influence model (U-shaped or W-shaped). Pick both up front, then lock them.

You do not need a custom attribution platform. You need one sourced model and one influence model, chosen up front and applied the same way every time. Here is the short version of the standard options.

Single-touch (sourced). One campaign gets 100% of the credit. First-touch credits the origin; last-touch credits the closer. For events, first-touch via Salesforce Primary Campaign Source is the usual sourced choice. Pro: simple, finance-legible. Con: ignores everything else that moved the deal.

U-shaped (multi-touch influence). Credit splits with the heaviest weight on the first touch and the lead-creation touch, the rest spread across the middle. Good when the event is often the moment a known prospect becomes a real opportunity.

W-shaped (multi-touch influence). Adds a third weighted touch at opportunity creation. Heavier model, better for longer enterprise cycles with more touches, more setup overhead.

The decision rule is mix, not maximalism. Pick the simplest sourced model your finance team trusts (almost always Primary Campaign Source) and one influence model that matches your sales cycle length. Short cycle, fewer touches: U-shaped is plenty. Long enterprise cycle: W-shaped earns its complexity. Running three models in parallel produces three numbers and an argument, not clarity.

Whatever you pick, the operational requirement is the same: the event has to be written onto the opportunity as a campaign with the event name, at the time of capture, so both models have something to read. Salesforce plus Campaign Influence plus a disciplined Primary Campaign Source field gets most teams most of the way. HubSpot's revenue attribution reports plus campaign associations do the equivalent. The model is the easy part. The campaign-member mapping you set before the show is the part that actually decides whether any of it works.

What are realistic trade-show conversion benchmarks?

Trade shows hit 24% lead-to-MQL per First Page Sage, conferences 28%, executive events 54%. By vertical, expect 13-16% MQL-to-SQL.

Benchmarks are for sanity-checking your own funnel, not for copying into a board deck as your forecast. With that caveat, the cleanest by-channel cut comes from First Page Sage's lead-to-MQL benchmark by channel. Every number below is per First Page Sage, which is the dominant publisher for this data, so treat it as best-available-single-source rather than independently triangulated.

Lead-to-MQL conversion by channel, per First Page Sage:

Two things to read out of that table. First, the in-person band (executive events, conferences, trade shows) sits high relative to digital-only channels, with executive events near the top alongside referrals. Second, trade shows at 24% land below conferences but above the webinar floor at 19%, which fits how a trade show works: high volume, mixed intent, a lot of badges that were curiosity rather than buying. The funnel quality is good, the top of it is noisy. That is exactly why booth-level qualification matters for the ROI math.

Down-funnel, the rates depend heavily on your vertical. Per First Page Sage:

For fintech specifically, MQL-to-SQL is contested between sources, so use a range of roughly 11-19% rather than a single point. Stack these against your own pipeline. If your trade-show MQL-to-SQL is running at half the benchmark, the problem is usually qualification at the booth or follow-up speed, not the show. If your lead-to-MQL is far below 24%, you are likely scanning everything that walks past and calling it a lead. Benchmarks tell you which stage is leaking, not what your forecast should be.

What does a trade-show lead cost?

Plan on roughly $100-300 per trade-show lead, varies by show and booth size. Full-funnel blended cost per lead runs higher and varies by vertical.

Be careful here, because this is one of the most mis-cited numbers in event marketing. The honest direct cost-per-trade-show-lead figure is a range, not a point. Listicles quote three different "per CEIR" numbers that do not reconcile and do not link to an actual report, so anchoring on a single precise dollar figure is false precision. The defensible framing is roughly $100-300 per lead, varying by show and booth size. If a vendor gives you one exact number, ask for the source.

That direct number is also not the same as your blended cost per lead, and the two get conflated constantly. Blended CPL is full-funnel: it folds paid plus organic plus content plus events into one number per industry. It runs several times higher than a single-channel cost and differs hard by vertical. Per First Page Sage's blended B2B cost-per-lead benchmark:

Two rules for using that table. It is full-funnel blended, per First Page Sage, so do not blend it with a Google-Ads-only cost-per-lead number from a different source as though they measure the same thing. They do not. And do not read it as "what a trade-show lead costs you." It is the all-channel average for the vertical, useful as a reference point for whether your event leads are cheap or expensive relative to your other channels.

The cost most teams forget is the scanner itself. Organizer lead-retrieval rentals run $400-700 per device per event, reaching about $735 onsite at the largest shows, and organizers increasingly charge nearly as much again for API or CRM-integration access on top. A five-person booth renting handhelds across ten shows a year is $25,000 to $35,000 in scanner rentals alone, before any software. That line item belongs in your all-in spend, and it is the part of the cost math that has nothing to do with lead quality and everything to do with how the show is priced. The full breakdown is in the hidden cost of organizer badge scanners.

How many leads should you expect to spread that cost across? The rough industry range is 50 to 200 leads per event depending on booth size, with no authoritative point average. Use your own last three shows as the real baseline. Booth size, floor traffic, and how aggressively your reps scan move that number more than any benchmark will.

Why do most teams underreport event ROI?

Only 6% of marketers convert show leads extremely well, 19% have no strategy to act on leads, and under 10% of exhibitors qualify leads at the booth.

The numbers tell the story plainly. Per the CMO Council and E2MA's Customer Attainment From Event Engagement survey of 260-plus brand marketers, just 6% of marketers say their company does extremely well at converting tradeshow leads into customer business. Same report: 19% lack a strategy to act on the leads they gather, and 45% struggle to make the case for attending at all. That last one is the ROI-measurement gap stated directly. Nearly half of exhibitors cannot build the argument for the budget they are spending.

It gets more concrete upstream. Per CEIR, citing major lead-retrieval vendors Experient and CompuSystems, less than 10% of exhibitors use customer qualifiers at the booth. Read that next to the attribution problem and the cause is obvious. If you do not qualify the lead at the point of capture, you have no clean way to separate a real buyer from a badge that wanted a tote bag. Three months later, when you try to attribute pipeline, the source data is missing. The model has nothing to work with, so the ROI report becomes guesswork, and guesswork is what makes a CMO distrust the channel.

You have probably seen the famous claim that "80% of trade-show leads are never followed up." Be skeptical of it. That figure is poorly sourced. The trail forks several incompatible ways, and the organization it is most often pinned to, CEIR, publicly says it is not from any research they generated. The properly sourced version of the same gap is the 6% number above, which survives a source check. Use that one. The reason this matters for ROI is that bad stats lead to bad fixes. Teams chase a follow-up-speed problem they cannot verify instead of fixing the qualification-and-attribution problem the data actually shows.

The underreporting is not because events do not produce. It is because the measurement chain breaks at the booth (no qualification), again at the handoff (no clean source data into CRM), and a third time at the model (no agreed attribution rules). Fix those three and the ROI number stops being a shrug.

How does Tendro fit event ROI measurement?

Tendro is the capture layer. It writes clean, qualified, event-tagged leads into your CRM at the booth so your attribution model has source data.

Disclosure: I build Tendro. Filter accordingly.

Every section above lands on the same root cause. Attribution is a data problem before it is a modeling problem. The slickest U-shaped or W-shaped model produces garbage if the source records are dirty, untagged, or arrived two days late without notes. The fix is upstream, at the moment of capture, which is the part most teams treat as an afterthought.

Here is where Tendro sits in the ROI chain, mapped to the model in this piece.

Capture clean source data. Scan any badge, business card, QR code, or note. The record is OCR'd and enriched at the booth, offline, so you are not retyping a CSV three days later or losing rows to a broken import. Clean inputs are the precondition for inputs 2 through 7 above meaning anything.

Qualify at the point of scan. Hot, warm, cold, the ask, the next step, voice or text notes attached to the lead. This is the step that under 10% of exhibitors actually do, and it is what lets you separate sourced from influenced pipeline honestly later. Without it, every badge looks the same in the CRM.

Tag the event and sync with attribution intact. The lead writes into Salesforce, HubSpot, Pardot, Marketo, and other destinations in seconds, attached to the event campaign, with the Primary Campaign Source set. That is the single mapping that decides whether your sourced-pipeline report works. Set it at capture, not in a cleanup pass that never happens.

What Tendro does not do: it is not your attribution model and not your CRM. It does not decide whether you run U-shaped or W-shaped, and it does not replace Salesforce Campaign Influence. It feeds them. It makes any attribution model executable by giving it clean, qualified, event-tagged source data captured at the booth. The model is your call. The data quality the model depends on is the part Tendro owns.

The case for measuring event ROI properly is independent of which capture tool you pick. If you want to see how the same data layer drives the CRM-sync side, the event leads to CRM hub goes deep on it, and the full guide is the event lead capture pillar. If you are weighing tools, the alternatives breakdown lays out the options.