The Hallway Track Was Always the Product

The Hallway Track Was Always the Product. The real problem with declining in-person event attendance is that we are going to lose if we do not fix the economics and the demographics.

At SQLBits this year, I chaired a panel about whether AI will replace us. Andy Cutler and I had spent several weeks thinking about how to frame it, based on a conversation we had had about consulting. I met Andy through the community many years ago, and he is now one of my good pals. In preparation, we had the arguments lined up, the counterpoints ready, and the data to hand. And then a hundred or more people walked into the auditorium, and the conversation went somewhere none of us had planned.

That is not a complaint. That is the point. SQLBits had specifically asked for panel discussions for this year’s event, and I think Brent may have called it the “productionisation” of the hallway track, I’d not thought about it like this before. However, having now hosted an SQBits panel discussion that landed with a large (100+), engaged audience, I think it is a great way to describe them.

Afterwards, someone posted a comment on LinkedIn that I haven’t managed to shake off. Chris Webb, commenting on a LinkedIn thread where Brent Ozar had shared one of my blog posts, Will AI Replace Us?, asked, almost in passing, whether I would have written that post if I had not gone to SQLBits and had the in-person connection with the audience. Honestly? No. I would not have had the idea, framed it the way I did, or had the confidence to publish it, without having stood on the stage and then in a corridor in Wales afterwards talking to people who had different experiences around the problem and, in some cases, knew far more about it than I did.

That single question is, I think, the answer to the entire debate about declining in-person event attendance. And I want to spend the rest of this post explaining why, because I think the community is sleepwalking toward something that will be expensive to rebuild once it is gone.

The conversation that is already happening

There has been a lot of talk about attendance at in-person events declining. Steve Jones published an honest update on the state of SQL Saturday in August last year, acknowledging that the number of events may stagnate or decline. He pointed at the obvious pressures: organisers are stretched, space is harder to find, and supportive venues that used to host events for free have closed. Then Brent Ozar picked the thread up and asked why people are not going to local and regional in-person events anymore. He listed around ten reasons, including  COVID hangovers, Zoom fatigue, a slowing release cadence in SQL Server, a generation of data professionals who entered the workforce remotely and never saw the value of being in the same room as their peers. They learned to do their job from their dining room table or bedroom. They survived — why would they give up their Saturday to attend a training event or conference?

Both posts are worth reading in full. Both make excellent points. And both left me with the same uncomfortable feeling, because I recognised myself somewhere in those lists.

This is a global problem. SQL Saturdays in the US, Data Saturdays across Europe,  I can only think of two data platform conference events that offer community days in the UK: SQLBits and Data Scotland, and I would consider both of those to be national or international events. Granted, regional user groups in the UK , at least the ones I have attended in person, are well attended. Which might expalin why there are no regional confercnence in th UK. The regional events it seems, all of them, are feeling the same structural pressures. The attendance trends are similar. The venue economics are similar. The sponsor conversations are the same. What I want to do here is go a level deeper than diagnosis, because the diagnosis has already been written (See Steve and Brent’s posts). This is my honest opinion of what is broken and what I believe it would actually take to fix it.

The lazy economics of staying home

From a pure cost perspective, attending a regional in-person event is a difficult business case in 2026. You lose a day of billable time. You pay for travel and often accommodation (That’s just me as a business owner. If you are an employee and funding this yourself its an even more difficult justification). The marginal cost of learning the same technical content has fallen through the floor. I can sit at my desk, watch a Brent Ozar recording over lunch, fire up a Microsoft Learn module between calls, or ask Copilot to summarise a topic in three minutes. If learning were the only reason we went to events, the case for showing up in person these days would be almost impossible to defend.

But learning was never really the only reason for attending. We just told ourselves it was, because it was the easiest thing to put on an expense form.

Is the hallway track the product we should build events around?

Brent makes this point near the end of his list, and I think it is the most important observation in the entire piece. We have never been good at marketing the hallway track. The casual conversations outside the session rooms. The chance encounter with someone doing something interesting at a different company. The five-minute exchange after a session that turns into an hour in the bar and a real working relationship six months later.

That is the product that might help reintroduce people to events. The sessions are the excuse to get everyone in the same building. The panel discussions is  the productisation of the hallway track. The panel discussion done well, will get a significant number of people sharing thoughts and ideas on the same topic. Probably chat about it afterwards, maybe discuss over a beer,  maybe even go away and write blogs and vlogs about the discussion so they can publish their own thoughts.

I will be honest: I do not close much consulting work directly from speaking or sponsoring data platform community events like these. The commercial return for me is longer, slower, and harder to attribute. But the marketing value, the relationships, the ideas, and the credibility that come from being seen in the room are real, and they compound over time in ways that LinkedIn impressions do not. Chris Webb’s question crystallises this perfectly. This post and my SQL Bits panel discussion posts exist because the conversation exists – including the panel discussion.

What the internet does brilliantly is breadth. What in-person does that the internet cannot replicate is depth, nuance and connection. It is the small disagreements that sharpen your thinking, the social pressure to actually finish the half-formed argument in your head rather than letting it die in your drafts folder, and the energy of a hundred people collectively wrestling with a problem in the same room.

You cannot stream that. You can broadcast it. You cannot replicate it. Whether it is the answer or not remains to be seen, but we need to start thinking differently about how we keep these alive.

The sponsor ROI problem nobody wants to say out loud

Here is the part of this conversation that might make some people uncomfortable nad perhaps some will disagree.  I want to say it plainly and state this based on my own experience only as a sponsor, but I know others have experienced the same, and if it is not addressed, it will have a significant impact.

Why does a sponsor sponsor a regional event? Let us not pretend it is community goodwill. There is some of that, particularly with long-standing vendors who have genuine community ties, and there are several. But in the main, sponsors sponsor because they need to justify a line item to a marketing director. That means measurable return.

Now look at what they actually receive for a gold package at a regional Saturday event. A logo on the website, maybe a link to their website. A pull-up banner in the lobby. A stand near the coffee. Maybe a lunchtime talk slot, which is a graveyard slot if ever there was one. And the opt-in attendee list at the end. The list is the reason I went down this route, sponsoring events myself.

Here is the problem with that list. Regional event attendees are, in the main, practitioners who want to learn. They are exactly the right audience if you are trying to influence a product’s end users. They are exactly the wrong audience if you are trying to close a deal, because they are almost never the decision-makers on procurement. The DBA in the room is not the person signing the contract for the monitoring tool. The data engineer is not the person approving the Fabric licence spend.

On top of that, attendees are increasingly reluctant to hand over an email address. Those who do tend to drop their card in the bowl for a raffle prize, which means the sponsor ends up with thirty to fifty addresses from people who wanted a Yeti cup (When I sponsored, it was an iPad. Times have changed), not a sales call on Monday morning.

The sponsor’s measurable return looks something like this: a few hundred pounds spent on logo placement, thirty raffle-quality leads, of which maybe five represent a meaningful conversation, none of them decision-makers. The sales team books two follow-up calls. Neither converts. The sponsorship line item gets quietly reviewed at year’s end and reallocated to an international event with a different audience, an analyst report, or a LinkedIn campaign that can be measured to the click.

I am not blaming the sponsors. The maths is the maths. But as the number of sponsors decreases, so does the ability to run the events. We need to stop selling a sponsorship proposition that only delivers for recruiters and has not delivered for other types of businesses for a long time, and then be surprised when renewals dry up.

What a different sponsorship model could look like

I am not expecting everyone to agree with this, and I am sure there will be counterarguments – they might be right! I’m thinking as I write.  But I will say it anyway. If the hallway track is the real product, then the sponsorship model needs to be built around access to that product, not around logo placement adjacent to it.

Here is the idea I keep coming back to, and it is just an idea, it’s not tested, not peer reviewed. Instead of giving sponsors a stand and a lunchtime slot (they can still have the stand), give them a meaningful seat in the formats that actually generate value. The panel discussion. The moderated roundtable. The structured debate where the audience is genuinely part of the conversation. These formats work because the value comes from the collision of perspectives, not from any single person’s delivery. A sponsor who is seen as a credible voice in a real conversation about a real problem is infinitely more valuable to their sales pipeline than a sponsor who is seen as the person who paid for the lanyards.

As an example, imagine a panel discussion around monitoring SQL Server, with an audience full of people interested in monitoring – everyone is interested in monitoring, and you have community folk like Erik Darling or Marcin Gminski. People who give their tools away for free, open-source, and talk with sponsors like Idera and Redgate, who are also there. That gives the sponsor a chance to engage with the audience in a meaningful way. And I do not mean a sponsored slot  nobody goes to those, I mean  sponsors integrated into a value-add discussion. Some sponsors actually do this themselves, and I suspect they are the ones that get the best ROI from their sponsorship. There are many that don’t do this.

Practically, this might look like: a roundtable with eight to twelve practitioners that include a few vendor representatives who contribute rather than pitch, with the sponsor getting access to the notes and the relationship rather than a badge scan. Or a sponsored “problem clinic” format where practitioners bring real challenges and vendors are one of several voices offering perspective.

The key shift is from passive visibility to active participation. Sponsors stop being furniture and start being participants in a valuable discussion. That changes their ROI conversation entirely, because they can now point at relationships and conversations rather than impressions and leads.

This also changes what the event is selling. Instead of “come and watch twenty sessions,” it becomes “come and be part of twenty conversations.” That is a harder thing to replicate on a livestream. That is a reason to get on a train or in your car.

The generational pipeline problem

There is a demographic argument underneath the economics argument.

The people who built the regional event culture, the organisers, the regular speakers, the veterans who turned up to every SQL Saturday and stayed for the hallway conversations, are, in the main, getting older. Some are retiring. Some have decided, entirely reasonably, that they have done their time and are not leaving the house on a Saturday unless someone is paying them to. Brent mentioned this in his post: the fifty-something DBA who has earned the right to coast and is taking it. I understand that completely. But when that person stops coming, they do not just take their technical knowledge with them. They take the social infrastructure. The introductions. The institutional memory. The quiet mentoring that happened in corridors that never got written up anywhere.

The question is whether they are being replaced. And from what I can see, the answer is: not in the community, even if they are being replaced in the workforce.

There are younger data professionals entering the industry. Although if the industry news is to be believed, it is not at the same volume because,  you know — AI can do that. But people are entering the industry. They are technically capable, often more so than the generation before them in certain areas. But they came in remotely, learned from YouTube and Microsoft Learn and well-prompted conversations with AI tools, and many of them have never been in the same room as a hundred of their peers. The community events that might have socialised them into the network never formed part of their professional experience.

This is not their fault. It is a structural gap that opened up during COVID and has not closed.

There are organisations trying to address part of this. New Stars of Data has done genuinely good work in finding and encouraging new speakers, giving people a platform earlier in their careers than they might otherwise have had. That matters, and it deserves support. But the speaker pipeline and the attendee pipeline are different problems. Can you grow the number of new voices willing to stand on a stage while the room in front of them continues to thin out?

The honest version of the diversity and inclusion challenge for regional events is not just about who is on the programme. It is about whether the next generation of data professionals sees any reason to be in the room at all. And right now, I am not sure we are making a compelling enough case.

Part of this might be AI. If you entered the data industry in the last three years, you did so knowing that the tools you are learning are being automated around you. The career calculus looks different. The urgency to build a professional network through community events may feel less pressing when you are not yet sure what shape your career is going to take. That is speculative, but I think it is worth mentioning.

What I am more confident about is the practical fix, and it is similar to what I argued in the sponsorship section. If we want younger practitioners in the room, the format has to give them something they cannot get online. Not a session they could watch on YouTube. A conversation they could only have in person. Mentoring roundtables. Early-career speaker tracks with genuine support rather than a slot and good luck. Structured introductions between junior practitioners and senior ones who are explicitly there to meet them. New Stars of Data is pointing in the right direction. The regional events perhaps need to follow.

Data Wales: a case study in what is actually broken

I want to ground this in something specific, because the abstract version of the problem is easy to nod along to and then ignore.

For the last couple of years, a small group of us (Justin Bird and I, in the main) have been trying to launch Data Wales (Comng Soon?) a Saturday-style community event in South Wales, in the same spirit as SQL Saturday and the regional Data Saturdays that used to dot the UK calendar. They no longer exist in the UK. There are some pretty active user groups that run monthly, and they still seem quite active and well attended, at least the ones I have attended in person, and perhaps that explains why there are no regional conferences — people attend those instead nad perhaps there is no room for other data saturday community events in the UK?. Anyway, I digress.

The appetite for our event has never been the blocker. Every time the idea surfaces, people say they would come, they would speak, and they would sponsor at a modest level.

The venue has been the blocker. Every single time.

We came close with a local university. We had a champion on the inside, a route to a space that would have cost us next to nothing. Then the people changed. The champion moved on. The conversation had to start again with people who had no context and no reason to care. That is exactly what Steve Jones was describing: finding a venue is a skill that requires time spent building institutional relationships, and those relationships are fragile.

Without a venue, the financial case collapses. A Saturday event for 150 to 200 people needs rooms, catering, and someone to unlock the doors at the weekend. At commercial rates, you are looking at a five-figure spend before you have printed a single lanyard. The community sponsorship money to cover that does not exist at the scale that regional events need, especially for new events. You need to run it once to prove the audience exists, but you cannot run it once without the funding to cover the venue (£10k at commercial rates). Chicken, meet egg.

I am being open about this because it is more useful than another abstract think piece. If anyone reading this has a contact at a Welsh university, a tech employer with a large enough office anywhere along the M4 corridor  – Cardiff, Newport, Swansea, anywhere in between, or a venue partner who would see this as a community investment rather than a commercial booking, I would genuinely like to hear from you. The community will turn up. The speakers will turn up. We just need a door to open.

Should we stop making these events free?

There is one more funding lever that the community has been reluctant to pull, and I think it is time to have the honest conversation about it.

A SQL/Data Saturday I attended and spoke at this year (2026) had around 300 registrations. The typical attrition rate for a free event,  people who register and do not show(you know who you are),  will vary by event, but at this event, it usually sits at around twenty per cent. This year it was closer to fifty. The organisers were, understandably, disappointed. The rooms were half empty. The catering had been ordered for three hundred. The sponsors had been sold an audience of three hundred.

This is not an isolated incident. It is a structural feature of free registration, and it is getting worse.

When something costs nothing, the commitment to attend costs nothing. Clicking register on Eventbrite takes four seconds and carries zero social or financial obligation. Life gets in the way on the morning of the event, and the calculation is simple: staying home costs nothing, going costs time and travel. The free ticket does not tip the balance.

A small registration fee changes that calculation entirely. Not because fifty pounds is a meaningful financial barrier for most working professionals, but because paying fifty pounds is a decision rather than a click. It creates skin in the game. It makes the no-show feel like a loss rather than a neutral outcome. And practically speaking, fifty pounds a head with two hundred attendees is ten thousand pounds toward venue costs, which, depending on the city, could cover a significant portion, if not all,  of what organisers currently have to raise entirely through sponsorship. It might help Data Wales with its chicken-and-egg problem.

The obvious objection is that charging will reduce registrations and price out people earlier in their careers. The answer, I think, is a tiered model: a nominal fee for employed professionals, a reduced or waived fee for students and career changers, and a complimentary pass mechanism that sponsors can offer to their own communities. Eventbrite handles this natively. It is not a complicated system to administer.

The deeper concern –  that “free” is part of the identity of SQL Saturday and Data Saturday, and that charging changes what the event fundamentally is about,  is worth taking seriously. But I would rather have a paid event with two hundred committed people in the room than a free event with a hundred, half-empty tables, disappointed sponsors, and an organiser who burns out and does not come back next year.

The free model served the community well for a long time. It is not obvious that it is still serving it now.

So would I have written any of this without going to SQLBits?

No.

And I would probably have fewer interesting things to say, a thinner network, and a duller perspective on the work I do every day. In-person events add value.

The internet is a wonderful thing. I make my living on it. But it has never once given me an idea I could not have stolen from somewhere else. The in-person conversations have given me ideas I could not have had alone. That is not nostalgia. That is the commercial case for showing up.

The question is whether the events will still exist in five years for the next version of me to stumble into.

Chris Webb’s question has an answer. The harder question is what we are all prepared to do to make sure someone else can answer the same questions, in  the same way, a decade from now?

If you have thoughts on any of this,  the sponsorship model, the venue problem, the hallway track argument, or the Data Wales situation specifically,  I would genuinely like to hear them in the comments.

And if you want to read the posts that sparked this conversation, Brent Ozar’s post on why people are not attending regional events, Steve Jones’s on the state of SQL Saturday, and Chris Webb’s comments are worth finding on LinkedIn if you have not seen them.

Useful Links

AI Enablement Programme | Microsoft Copilot & AI Consulting