Sold by the Seat, Bought by the Room

Why the Model Had to Change

The first years were heady, and the company was not standing still. It listened, adapted and improved. The response was strong and we were being called in for the most spectacular installations in the market. Revenue was clocking steadily. We put marquee work into retail environments, building lobbies, control rooms, high-profile presentation venues and customer experience centres, and we had real visibility.

The problem was not the product or the numbers. It was the shape of the business underneath them.

Because the displays could be built to almost any size and geometry, the largest projects were also the most bespoke — and a bespoke installation is, in engineering terms, close to a new product every time. That does not scale. Neither did the rhythm. A custom project took three to six months to sell and another three to six to install, depending on site readiness. Revenue arrived in blips: infrequent, unpredictable, and large when it came.

The Cookie-Cutter Answer

The answer was the enterprise customer — the multinational with offices across dozens of countries and a standard template for how a meeting room is fitted out. That is a very large body of addressable demand which does not want anything customised at all.

So we built for it: a display in two standard sizes, one for mid-sized rooms and one for large ones, with the engineering already done and boxed. If a customer also wanted an experience centre or a retail installation, we were still delighted to customise. But the core of the business could now be a steady flow of standard orders and a backlog we could fulfil in a planned way.

That was the first half of the answer. The second half was not about the display at all.

Custom displays carry complex systems-integration requirements, and the least understood of them is content. To show something on a very large canvas you have to size for that canvas, and most buyers have never had to think about this. They commission an enormous wall and then discover their presentation occupies a modest rectangle in the middle of it. Selling big screens is therefore partly an education process: the useful question is not how large can we go but what do you intend to put on it, mapped and agreed before anything is ordered — because the person who has to live with the answer is not the technical buyer who specified it. It is whoever has to stand up in that room and use it.

Standard sizes solved how we would build and ship at scale. They did nothing about what appeared on the screen once we had it. That was a software problem.

What the Software Actually Did

The answer to that arrived as another start-up.

Their platform could treat a screen of any size as a single display canvas. It could take conventional inputs — a PC, a video player — and place them on that canvas as picture-in-picture at their native proportions. It could open Office documents directly: presentations, spreadsheets, PDFs, even web pages, from a source built into the display itself.

And it was touch. The wall behaved like an oversized tablet. You could position and play several pieces of content at once, pinch and zoom to resize them, annotate over the top, capture the annotated result and email it from the screen or keep it as a meeting note.

Then the part that made people stop walking.

Connected to a server, up to twenty-five endpoints could join a single session and mirror one another in real time. Everyone could annotate, share and work over the same content simultaneously, across locations. And the session persisted — you could leave it and return to exactly where you had been. A multinational with offices in New York, London, Dubai and Singapore could genuinely follow the sun: a team starts work in Mumbai, London picks it up and acts on it, New York reviews it in their own morning. The documents stayed on the display behind sign-in credentials, carrying the annotations and digital notes left by whoever had been there last.

Read in 2026 that is an unremarkable list. In 2014 it was not.

The Booth Behaved Like a Magnet

We showed it at InfoComm, and the stand pulled people off the aisle for three days. What made it work was that nobody needed to be talked into it. They put their hands on the wall and understood it in about four seconds.

One senior CTO, moving content around the canvas with both arms, said it made him feel like Tom Cruise in Minority Report. Another visitor called it the iPad for the enterprise. The second phrase was the more useful of the two, because it explained the category to people who had no reference point for it: not a bigger screen — a different kind of surface.

Then came the endorsement that mattered more than any of them.

The largest collaboration company in the world — one of the three I am about to describe as competition we could not match — became an early customer. They commissioned a large collaboration wall for the meeting room attached to their chairman's office, and went on to make the product a standard across their executive briefing centres.

Notice what they bought it for. Not to connect their own offices to one another; they had a platform for that, and it was one of the platforms we were up against. They bought it for the rooms where they show their own customers what the future looks like.

Even the company that owned the collaboration category valued this first as a room experience. I did not fully appreciate at the time how much that mattered.

The Case That Convinced Everyone

If this could be made to work, it was bigger than the display business.

Not one screen sold to one room, but thousands of users across dozens of offices — recurring subscription revenue, plus the hardware, plus service and support contracts on top. Investors understood that arithmetic immediately, and they were right to. It was a genuine opportunity and there was a real chance of being first into it.

So the software company was acquired and funded to turn the platform into a proper cloud product: a user licence, a platform licence, a web application, even an iPhone app.

We were suddenly in a different business.

The Compromise Nobody Named at the Time

There is a detail in that transition which looked minor and was not.

Our smallest display was 117 inches and the larger one 190. Those numbers were not chosen because research identified them as the sizes the market wanted. They were the sizes our existing modules and the engineering behind them produced. The product line was defined by what we could already build.

And an enterprise customer does not want screens only in its two largest rooms. It wants screens in all of them. So we introduced 65-, 85- and 98-inch LCDs — ordinary panels, priced nowhere near our own technology.

That was the first time the company compromised on the display. It diluted the display-technology fundamentals the founders had built the business on, and it did so quietly, as an accommodation to a go-to-market decision rather than as a stated change of direction.

Squeezed From Above and Below

Meanwhile the market moved underneath us in both directions at once.

Fine-pitch LED had been growing slowly for years and then hit its exponential stretch: cheaper every cycle, with pixel pitch getting finer each generation. Large fine-pitch LED was no longer distant from what we offered. At the other end, LCD bezels kept narrowing and panel performance kept improving. Our technology was being squeezed — LED pressing down from above, LCD pushing up from below.

And in the new business we had just entered, the competition was not other display manufacturers. In 2014, remote collaboration meant video conferencing, and content collaboration meant one person sharing their screen to a group. What we offered — several pieces of content on one canvas at once, democratically, visible in another room or on a laptop — was immediately recognised as a gap worth filling.

But the companies whose gap it was — happened to be Cisco, Microsoft and Zoom. Already the enterprise standard. Already installed, already subscribed to globally, already used daily by people who had no complaint about them. A small vendor with a better idea is not a match for that, and it is worth saying plainly that those platforms went on to absorb most of these capabilities. The features that made our booth a magnet in 2014 are ordinary now precisely because the giants eventually built them.

The chairman's wall sits inside that fact rather than against it. They bought the surface, because the surface was genuinely without peer. They did not need the network underneath it, because they already owned one.

We were unmatched at the part of the proposition that sold by the room, and outgunned at the part that was supposed to sell by the seat.

The Reorganisation, and the Two Regions That Escaped It

The company reorganised globally to match the new model: from hardware sold through a channel, to a solution sale of hardware and software together.

A customer success function was created worldwide and in every region, embedded alongside client teams to build use cases, design workflows and drive adoption. It was the right capability to add and it had an uncomfortable economic floor — it could only be afforded at sites carrying at least a few hundred user licences. Sales became account management: relationships, opening doors, closing, supported by sales engineers who did the solution work. A classic IT sales setup, in which the new generation of salespeople were generalists rather than people who understood the whole system and its technicalities.

The United States and Europe transitioned completely, and the capability they added was notably heavy. They also went at direct accounts, effectively starting from scratch.

New leadership and a new philosophy added uncertainty on top of a steep learning curve, and the most consequential casualty was the old channel. Building the new relationships was right. Losing the old ones would come back to bite, later.

Adoption was slow and difficult, and the results did not repay the effort. The company was no longer the light, nimble thing it had been as a start-up. It floundered under its own weight.

The Middle East and APAC were the only regions that came through unhurt or lightly marked. That was not luck, and it is the part of this story I would defend hardest.

We had six people in 2014 — a regional lead, two in sales, three in support. We went to eleven: one more salesperson, two more support engineers, a channel manager and a pre-sales specialist. We deliberately hired fewer people than head office had mandated.

Every one of those additions was chosen for a specific structural reason. The salesperson came from enterprise IT rather than AV, to carry the large regional accounts — and by design had the regional experience and connections to open new geographies. The pre-sales specialist structured the solutions and use cases inside proposals, and also carried customer success. The new support engineers were IT people, briefed to learn AV from the existing team while the AV engineers learned IT from them; they were staffed to cross-pollinate rather than to replace. The channel manager's job was to rebuild the partner programme and prepare the move to a distributor structure, a search that was already under way.

And critically, we did not abandon the sales motion or the partner relationships that had built the region.

What the Region Made Us Build

Two further adaptations were regional, and both are more instructive than they sound.

The first was data sovereignty. Cloud was well understood in the United States and hosting there raised no objection. In Europe it was a challenge, though global customers with large office estates could still take it. In the Middle East cloud was materially less acceptable, and on-premise versions were built to serve this region and Europe. Note what that means: the region did not reject the product, it rejected the delivery model — and the answer was engineering, not persuasion.

The second was cultural, and no amount of engineering would have found it from California.

The platform assumed that the person driving a session stands at the screen, touching it. Here, standing in front of a display and operating it by hand is not what the most senior person in the room does. He sits. So we offered a touch tablet on the desk that drove the wall remotely, and it was a considerable hit.

Nothing in a product roadmap surfaces that. It came from being in the room.

Where the Cross-Room Use Case Actually Turned Up

The multi-room capability did sell here. It sold for reasons nobody in California would have written down.

So the pattern is precise, and it is not the one the product was designed around. Multi-room, yes. Multi-site across geographies and time zones, almost never. The rooms that needed to see each other were in the same building, or on the same campus.

Follow-the-sun from Mumbai to London to New York was a magnificent capability. It was not a problem anybody here had.

Sold by the Seat, Bought by the Room

The region kept delivering its numbers. Software was under ten per cent of them.

The reason is not that customers disliked the software — it was our single biggest differentiator, and it won us work nobody else could have won. The reason is a mismatch between how the product was licensed and how the market bought it.

Licensing was per user, and it had to be: persistence and the ability to link personal storage — OneDrive, Google Drive and the rest — only make sense against an individual identity. But that was not our customers' workflow. Even with single sign-on and the integrations working properly, clients were unwilling to buy a licence for every employee for a use case their employees did not have.

So they bought user licences and assigned them to rooms.

Against an investment case built on thousands of seats per account, those are rounding errors. Which means the region was succeeding and the strategy was failing, in the same accounts, at the same time.

What Transferred, and What Held

The pivot did not cost us our accumulated position, which is the opposite of what usually happens.

The logos transferred. The reputation transferred. The applications changed, but most of our top-tier partners stayed — because they had always operated at the intersection of audiovisual and IT, and the market was moving in the direction they were already heading. We added new partners at that intersection rather than replacing the ones we had. Our profile improved, in fact: we were now addressing larger customers and bigger integrated projects.

The tender mechanic survived too. Large displays still carried a ticket size that required formal tendering, and the specification could now include things no competitor could deliver at all — multi-site and multi-room mirroring, session persistence, the software capabilities themselves. The lock-in from Part One did not weaken. It strengthened.

Two things that might have hurt did not.

Cash was never a problem. We sold subscriptions the way we sold hardware — annually in advance, or three years in advance against a heavy discount. Beside the price of the display it was a small enough line that nobody argued about it, and it settled naturally alongside the annual support renewal. The classic software-transition trap, where deferred revenue starves the business funding it, did not catch this region.

Support did not break. The display platform itself had not changed, so everything built in Part One kept working. Projects were order-to-install, so no partner was left holding inventory.

And one decision was made above us. We were never permitted to sell the software on its own. It always went out attached to a display. In the United States a software-only motion was explored, but the full system only came alive on a screen, and that was the corporate position. It limited both sides of the business — the software could not reach buyers who wanted only software, and the display could not shed the cost of carrying it.

That becomes considerably more consequential in Part Three.

What I Got Wrong

Four things, and the last one is the one I would change.

We chased the use case we wanted to be true. The most attractive model we built was a university campus with connected classrooms, a licence for every student and genuine collaboration across the institution. It was a beautiful proposition and we sold a pilot on it in Saudi Arabia — a handful of LCDs and fifty licences. It did not fly. The sector was well funded and budget was never the obstacle; the requirement needed more customisation than we could offer. We spent real time and effort on a model that was more compelling to us than it was workable for them.

I read the in-room pattern as regional. It was not. It was global, and it was not really a market problem at all — it was adoption and change management. It came down to habits, and to Teams, Zoom and OneDrive already being where people worked. Our story was so compelling that demonstrations converted at around ninety per cent on the use case, and where we lost it was almost always budget. But I have walked into rooms across Europe and the Middle East where a beautiful wall was being used as an ordinary screen with a laptop plugged into it — and everyone was perfectly happy with the screen and the room.

The wider truth is that even in-room collaboration is reserved for high-value occasions: a board meeting, a client presentation, a training session, a strategy workshop. Eighty or ninety per cent of meetings are somebody sharing a screen and talking about it, or making a video call. The company eventually learned this and shipped a quick-start mode — no sign-in, no account, plug in and start, one touch to place a video call, with the full stack available when it was actually needed. It was a hit. It did nothing for user-licence sales.

We carried the LCD line knowing what it cost us. It got us into smaller accounts, on the theory that they would graduate to a bigger wall later. It took exactly as long to close as any other deal. What it really did was keep the meter ticking for the sales team and the company.

And the one I would do differently. When we ran a demonstration, the person watching was already picturing himself in Minority Report. The solution was aspirational, and the technical buyer and the C-suite genuinely appreciated what it made possible. But the difficulty was never at the top of the organisation. It was in translating that aspiration into daily habit further down, and adoption was the entire game.

We could have done far more about that. We did not, because at the scale our customers bought — five licences, ten — a customer success programme could not be justified.

And we never attempted to sell adoption as a paid programme in its own right. That is the thing I would go back and change.

If You Are Entering a Region Now

The transferable lesson is not that pivots damage regional businesses. Ours came through with its reputation, its partners and its channel position intact, and in some respects stronger than before.

The lesson is narrower and harder. A subscription model is a bet on a specific customer behaviour, and that behaviour is not evenly distributed across markets. Ours was a bet on distributed teams working across offices and time zones. That behaviour was real in North America and largely real in Europe. Here, the same technology solved a different problem — segregated classrooms, crisis rooms down the corridor — and solved it well, but with five licences instead of five hundred.

That conversation is far easier to have before the targets are set than afterwards.

Where This Leaves the Region

By the end of this period we had a solution the market valued, a channel that had evolved with us, and a sales motion built around accounts rather than opportunities. What we did not have was reach — the ability to cover the IT and collaboration channel properly without simply adding headcount we had already decided not to add.

The next move was to find a distributor who already lived in that channel. And the next product was about to change the argument again.