Conference Research Without Software: The Manual Method and What It Costs

You booked the flight. You paid the registration. Nobody sent you a list of who else is going.

This is the manual method for working that out. It uses eight sources, all of them public or semi public, and no software. It is the method most teams already use, and for a lot of teams it is the correct one. It is also worth being honest about where it stops working, which is not where most people expect.

Why does no one hand you the attendee list?

Because the list is the product. Organisers sell access to the room through sponsorship, and Forrester's Q1 2026 State of B2B Events Survey found 51 percent of B2B event budgets now go to sponsoring third party events rather than hosting owned ones. The attendee list is what that money buys. Publishing it would devalue the thing they sell.

So the room has to be assembled rather than requested. That is annoying but not hard.

The eight sources

Ranked by signal per hour, not by volume. A verified list of 40 speakers is worth more than 400 maybes.

Source	What it gives you	Rough time	Reliability
Speaker list and agenda	Verified names, titles, companies of the most senior people in the room	1 hour	Very high
Sponsor and exhibitor list	Companies that committed budget to be there	1 hour	Very high
Event app attendee directory	Names, titles and companies of everyone who opted in	2 to 3 hours	High
LinkedIn event page and hashtag	Self declared attendees, with stated reasons for going	2 hours	Medium
Past editions of the event	Which companies return, and what they send	2 hours	Medium
Community Slack and WhatsApp	A small, warm slice already talking about the event	Varies	Medium
Ticket tiers and session sign ups	Who paid for the expensive room	1 hour	High
Your own CRM	Existing relationships and pipeline overlap	1 hour	Very high

Times are planning estimates for one event, not measured averages.

Start with the speaker list and the agenda. Always public, always the most senior slice, and an hour well spent. Pull every speaker, panellist and moderator into a sheet with company, title and session. Then read the session descriptions, which is the part people skip. Case study sessions name customers, which hands you a live vendor relationship. Panels show who is willing to be seen next to whom.

Then the sponsor and exhibitor list. This is the cleanest budget signal available for free. Cvent's trade show research puts the average total cost of an exhibit at 10,000 to 30,000 dollars per booth, so every logo represents committed spend in your category. Rank sponsors by tier, then check which are new against last year's page. New sponsors are entering the category and are usually the most open to a conversation.

The event app directory is the largest single source, and the price of entry is registering early rather than the week before. Bizzabo's 2026 State of Events Benchmark Report found 55 percent of attendees say the mobile event app can make or break their experience, which is why organisers keep investing in it and why opt in rates hold up. Two limits worth stating plainly: it is opt in, so it is never the full room, and it is semi public, so it belongs in your research and not in a sequencer.

LinkedIn is a small channel with a warm audience. Search the official hashtag sorted by recent. The value is rarely in the posts, it is in the comments under the "who else is going" ones, which function as a self assembled list with intent attached.

Past editions are more predictive than people expect. Conference audiences are sticky. The same sponsor tiers, the same job titles and often the same individuals return year after year, particularly at established events. Pull last year's speaker list, sponsor page, recap post and photo gallery. Sponsors who returned three years running are the safest bets on the page.

Community groups, ticket tiers and your CRM round it out. Workshops and paid side events usually cap at 20 to 40 people and circulate participant lists far more freely than the main event, and everyone in that room self selected into your topic. Your CRM goes last, cross referenced against everything else, to surface open opportunities and dormant accounts you did not realise were attending.

The part nobody warns you about

Every guide to this stops here, at the point where you have the names. That is the wrong place to stop, because sourcing is the easy half.

By the end of a determined day you have somewhere between several hundred and a few thousand rows. They are in four formats. The same company appears as "Acme", "Acme Inc" and "Acme Technologies". Titles are inconsistent across sources, because a Director of Partnerships at one company does the job a VP of BD does at another. Roughly a fifth of the rows are stale, because people changed jobs between the speaker page going up and the event happening.

Now do the actual work. Which of these companies fits your ICP? Which of the people are in a position to buy, rather than merely present? Which twenty are worth the two days you have on the floor?

That is where the week goes. Not the finding, the deciding.

What it converts to

Worth knowing what the work buys you. HockeyStack Labs analysed 2,642,337 deals across 198 B2B SaaS companies and found live event deals converted from created to qualified at 5.50 percent, against 4.82 percent for all other channels. Events beat the average, but the gap is not enormous, and it only exists when the room is worked deliberately. A conference attended without a target list converts like a conference attended without a target list.

Freeman's 2025 commerce research found 32 percent of attendees are the final decision maker on a purchase, with a further 68 percent holding some or significant influence. The people worth meeting are genuinely there. They are just mixed in with everyone else.

When the manual method is the right answer

For one or two events a year, it is fine. Better than fine. A day of sourcing and a couple of days of scoring is a reasonable price for a trip that costs thousands, and doing it yourself builds a feel for the room that no tool hands you. If you attend Money20/20 and nothing else, do it by hand.

It also keeps you honest. Reading session descriptions and sponsor tiers yourself teaches you things about a market that a ranked list never will.

Where it stops

It breaks on the third event, and it breaks for a specific reason: none of the work carries forward.

The dedupe you did for one event does not help with the next. The ICP judgements live in your head and in a spreadsheet nobody else opens. When a colleague attends the next one, they start from an empty sheet. Do this four times a year across a team and you are spending roughly a person month a year reassembling the same understanding of the same market.

Forrester found only 20 percent of B2B event teams currently use AI to define and build attendee lists, which means the manual method is still the industry default rather than the fallback. That is either reassuring or an opportunity, depending on how many events you attend.

The honest test is not whether you can do this yourself. You can, and this article is the method. It is whether you will do it again in eight weeks for the next event, and whether the version you do then will be any better than this one.

If you would rather not

Sideroom does the second half. You bring an event and your ICP, and it returns the companies and people ranked against it, with the deduping and enrichment already done.

The first half, the eight sources above, is genuinely worth understanding either way. Knowing where the information lives is what lets you judge whether a ranked list you did not build yourself is any good.

You can see what past editions of an event looked like with the free conference history tool, which takes a company name and needs nothing else. If your ICP is the part that is unsettled, start with defining who is worth your time, then turning a raw list into a ranked target list, and the wider question of who to meet at a conference.