New York, London, Dubai

The Third Leg

In late 2009 I met the founders of a Silicon Valley start-up with a new display technology, a product they believed was close to shipping, and a decision already taken. The company would build three commercial legs at once: New York for the Americas, London for Europe, Dubai for the Middle East.

I was expected to join in a hurry — the launch was already slated for February 2010, at ISE in Amsterdam.

By the time the show opened, production had slipped, so what actually stood on the stand were engineering samples and pre-production units rather than the finished product, behind a cordon, shown by invitation only.

My first day with the company was at that show. I joined officially the following month, in March 2010.

Two of those legs began with something. New York was the company — head office, engineering, product, and a home market everyone already understood how to sell into. London had a market a US board could recognise, and a person to run it.

Dubai had a thesis.

No entity. No employee. No office. No installed product in the region — and no proven customer base anywhere else in the world to borrow credibility from.

The company was launching globally from zero, and the Middle East was starting from zero inside that zero.

This is the impasse every vendor eventually reaches, and it is circular. You cannot justify a regional entity for a business that does not exist. You cannot build the business without someone in the region. Most companies resolve it by deferring — appoint a distributor, or run the market on quarterly visits, and revisit in eighteen months.

We resolved it differently, and not cleverly. The answer was a person before a structure: I joined full-time as a consultant, because there was no regional company through which I could be employed. That was a workaround. It was also the only thing available, and it set the sequence for everything that followed.

One structural point deserves naming, because it is usually reported the other way round. The Middle East was never carved out as a standalone territory. It sat inside International, clubbed with Europe — a sub-territory inside somebody else's number, which is precisely the arrangement vendors are normally warned against.

It made less difference than you would expect. What mattered was that one person was accountable for the region and physically in it. Structure follows that; it does not substitute for it. Plenty of companies have created a standalone Middle East region on a slide and staffed it with a quarterly flight.

Fifteen Months Before Anyone Could See It

The product was announced at InfoComm USA in June 2010, but it was not commercially available until early 2011, and a demonstration unit did not reach the region until June of that year. From arriving in March 2010, I spent roughly a year in a market with nothing to sell, and fifteen months before a customer could physically look at the thing.

That is long enough to destroy a market entry, and the two usual ways are avoidable. The vendor goes quiet, because there is nothing to announce. Or the vendor stays loud, burns its introductions on a product that keeps slipping, and reaches launch having already spent its credibility.

There is a third option, and it is the only period in a market entry that is genuinely free. Before a product exists, nobody expects a number from you. That is the one window in which it is possible to ask basic questions in front of senior people without looking like you are wasting their time — because you are not yet asking them to buy anything.

The starting point was the systems-integration community. Partners already knew which projects were moving, which consultants were designing them, and which end users were dissatisfied with what was available. We introduced the concept and asked where it could fit, which projects were early enough to influence, and who would ultimately shape the decision.

Where partners had the relationship, they introduced us to the customer or the consultant. Where a project was visible and no introduction existed, we went directly. Partners led us towards projects; projects led us towards partners.

That two-way movement exposed the distinction the rest of this article rests on.

A route to market answers how the product reaches the customer. A route to business answers how the customer reaches a decision — who identifies the problem, who designs the solution, who writes the specification, who bids, who carries the commercial risk, who installs, who supports, and who remains accountable when something changes.

The partner is one part of that chain. It is not the chain.

What the Customer Was Being Asked to Absorb

The technology was new: laser light exciting phosphors across modular tiles, assembled into large and almost seamless configurations. Every alternative then available carried a compromise — projection needed controlled lighting and consumables, LCD walls had heavy bezels, plasma was heavy and prone to burn-in, and fine-pitch LED had not yet arrived.

Our product addressed most of those limitations. It was also substantially more expensive. So technical interest was not the test. The test was whether applications existed in which the difference mattered enough that price stopped deciding — and broadcast studios, command environments, control rooms and high-end institutional spaces began to supply the answer.

But look at what the first customer was being asked to accept, all at once: an unproven technology, an unknown American vendor, no meaningful reference base anywhere on earth, and a premium price.

Nothing on that list is specific to hardware. Any vendor entering the Gulf without a regional reference asks a customer to absorb the same four risks, whether the product is a display or a platform. The first sale depends less on the product than on the credibility of the people behind it and the support model they can actually demonstrate.

A Screen in a Rented Office

In June 2011 a small version of the display finally reached Dubai — fifteen months after I had started selling the idea of it.

There was no showroom waiting for it. We installed it in our temporary business-centre office — a Regus suite, which is not where a customer expects to evaluate a seven-figure infrastructure decision. For the first time customers could see what the animations had tried to explain: brightness, viewing angle, the narrow joins between modules, and how the same building blocks could produce much larger walls, unusual shapes and curved screens.

The problem was getting the right people in front of it.

We could not expect senior decision-makers to visit an unknown vendor's temporary office. So for every demonstration that mattered, we dismantled the display, packed the modules, transported them to the customer's site, rebuilt the wall, calibrated it, ran the meeting, and reversed the entire process. A full cycle ran to four or five days, staffed by one technical employee plus partner resources and outsourced labour.

That last figure is worth sitting with, because the obvious explanation for it is only half right.

For a visual technology, seeing is believing. But the conversion rate was not high because the product was persuasive in the room. It was high because of what had to happen before anyone got into the room. The customer had to give up space, arrange access and assemble the right people on a fixed date. The partner had to put its own labour and credibility behind the request. And we had to spend fifteen hundred dollars and a working week. When all three agreed to absorb that, the opportunity was real — and when any one of them hesitated, we had learned something more useful than the demonstration would have told us.

Cheap demonstrations qualify nothing. This one was the most reliable forecasting instrument we had.

Nor was a building the answer. A permanent demonstration centre, built before the market is proved, is a bet; a mobile unit is a test. We were not yet entitled to the bet — and the money we did not spend on a showroom in 2011 was still there when we needed it for people the following year.

Spend on evidence before you spend on appearance.

Two Days Became a Week

The first significant visible opportunity surfaced almost accidentally, in a conversation with a systems-integration partner. A television programme in the UAE was due to go live in the first week of December and needed a new set design; the partner asked whether we could bring the display into the studio for the broadcaster to evaluate.

We said two days.

Consider what was being risked. The product was weeks old and we had no reference installation anywhere in the world. This would not be a showroom walk-through — the screen would be tested against broadcast cameras, studio lighting, colour temperature, white balance, clock synchronisation and the wider production chain, in the conditions under which it would appear on air. And it would be tested in front of the partner who had opened the door and the consultant advising the broadcaster.

If it failed there, every integrator in the region would know inside a month. In a market this concentrated, a first demonstration is not a sales meeting. It is a public examination.

The two days became a full week.

The technology was approved. The broadcaster then reduced the initial screen size to fit the budget available — and because the system was modular, the smaller configuration did not compromise the wider design. The programme went live in December 2011. A year later, when more budget was released, the broadcaster expanded the screen by adding modules, and the surrounding set construction and back-end integration did not have to be touched.

That first project did four jobs at once. It proved the technology under real operating conditions. It created a visible regional reference where none had existed. It demonstrated the commercial logic of modular expansion. And it prequalified us for the broadcaster's much larger news-studio upgrade.

That next project became our first order above one million dollars.

Specification, and the Moment the Channel Reverses

Large projects were awarded through formal tenders. Waiting for a tender to appear would have left us competing as an unfamiliar and expensive product against established alternatives, on someone else's terms.

So the work that mattered happened earlier. Once an opportunity was qualified, we engaged the end user and the consultant, demonstrated the technology, and helped develop the performance requirements — supplying draft specification language they could adapt. The resulting tender might name the technology directly, or specify characteristics only it could deliver at the time: extremely narrow joins, lifetime uniformity, modular scale, no consumable light source.

If you sell software, this mechanic has not changed; only the artefact has. The equivalent of the specification is the requirements document, the reference architecture, or the security and integration criteria a consultant or systems integrator writes months before an RFP is issued. Whoever is in the room when that document is drafted has already won most of the deal.

When the tender was released, several integrators would approach us. This is where the partner relationship became genuinely two-way. We assessed whether the partner had customer access, technical capability, commercial commitment and willingness to invest in training, tools and spares. The partner assessed whether we could help them win a profitable project and support it properly afterwards.

And the most recognisable integrator was not automatically the right one. Early in a market entry, a smaller specialist has more reason to prioritise a new product than a large house carrying twenty established lines on which it already makes its margin.

Once customers began specifying us, the dynamic reversed. Integrators that had shown little interest now needed us to stay competitive. Demand from the customer created pull through the channel — the only reliable way to get a partner network to work, and the opposite of how most vendors try to build one.

The Entity Followed the Work

The regional entity was established in September 2011 — after the first display had arrived, after the market had been mapped, and weeks before the first project went to contract.

Its purpose was not invoicing. Quotations, billing and collections continued through the US entity, sparing the regional operation a finance back office it did not yet need. The entity existed to employ people locally and build permanent support capability.

Until then, an engineer from the company's R&D team in India had been flying in and out of Dubai to support demonstrations and technical work. He became the first local employee.

That sequence was deliberate. For an expensive and unproven technology, technical support is not an after-sales function to be added once volume justifies it. It is part of the sale. A poorly installed or neglected early project becomes evidence against every future one.

The team eventually reached about ten people: regional leadership, three salespeople, one pre-sales specialist, and five support engineers.

Half the team was support. In a market with no references, that ratio is not an operational choice. It is the sales strategy.

Proof That It Travelled

The television studio created public visibility. What followed over the next three years is the part worth studying, because the order of it was not accidental.

Year one produced four references across three countries. The Dubai studio went live. A government museum in Oman. A retail showroom in Kuwait. And the follow-on at the same broadcaster — the news-studio upgrade that became our first order above a million dollars.

Year two moved into a new sector and a new market. Defence installations in the UAE, which brought a category of customer that does not buy on specification alone and does not appear in any case study. And our first project in Egypt: another television studio, the same use case that had opened Dubai, now travelling on its own reference rather than on ours.

Year three was the largest, and the type of project changed. Qatar and Abu Dhabi — control rooms, auditoriums and meeting-room environments. In Qatar a single university lecture hall led to the product being specified across an entire new campus: five further lecture halls and an auditorium, from one room that behaved.

Read that as a shape rather than a list. One visible installation, in one city, in one use case. Then the same use case in an adjacent country, carried by the reference rather than by us. Then adjacent sectors. And only in year three the categories with the largest budgets — control rooms, defence, institutional infrastructure — which are also the categories that take longest to qualify for and which no new vendor can enter first.

That order cannot be compressed by spending more. A control-room buyer will not be your first customer, whatever your discount. The same pattern is familiar to anyone selling software, and it moves faster in the Gulf than in most markets, because the buying community is small and it talks.

Many of the largest projects could never be publicly named — installed in sensitive facilities where photographs and case studies were impossible. But the trade knew. Consultants knew what had been designed, integrators knew what had been tendered and delivered, and customers in adjacent sectors knew which technologies were turning up in major projects.

The visible installations created recognition. The confidential ones created authority.

Winning Beyond Your Size

Winning multimillion-dollar infrastructure projects was commercially important. Delivering them was a different problem, and it was the one capable of ending the company.

Some contracts required factory acceptance, inspection and delivery as a single coordinated lot. Manufacturing, freight, storage, inspection, documentation, delivery and installation all had to align on one date — and that date belonged to a construction site nobody at the vendor controlled.

Site not ready is not an exception. It is the base case.

Material would be built to the contracted schedule, packed, labelled and standing ready for dispatch, and the building would not be. Containers then have to go somewhere. Partners rented temporary facilities to stage complete inventories so they could be held, inspected, tagged and released before shipping. None of that sat in anyone's budget, and the cost did not stay with the customer.

Then the money. The company was a US start-up accustomed to selling on unsecured credit backed by credit insurance — sound enough at a few hundred thousand dollars. At several million, against a regional integrator, on a project tied to a government completion date, it is not. We had to introduce letters of credit and comparable instruments, and then do the harder part: persuade a head office that had never needed them that the exposure justified the friction. The installations themselves had to be staffed by experienced teams flown in from the United States and Europe, because nobody in the region had yet worked at that scale on this product.

Each of those is survivable alone. What makes them dangerous is that they arrive together, on the largest order the company has ever taken, in the market with the least infrastructure behind it. A young company can win beyond its apparent scale. It can only deliver beyond its scale if people who have handled projects of that size are involved before the order is signed rather than after.

Above a certain value, winning stops being a sales event and becomes a balance-sheet event. The question is no longer whether you can sell it, but whether the company can absorb the working capital, the storage exposure, the receivable and the delivery risk if one dependency slips — and whether anyone has calculated what happens if it slips badly.

We made it happen. But what made that possible was not enthusiasm. At that size, an order you cannot deliver does not cost you a project. It costs you the company.

What We Got Wrong

An entry story told as a sequence of wins is not an entry story.

We lost a great many opportunities. On price, against alternatives that were adequate and cheaper. On specification, where we arrived after the requirements had already closed. On credibility, where an unknown American start-up was simply the safer thing to say no to. And frequently on some combination of the three, which is the usual shape of a loss and the hardest kind to learn from.

The learning was not to argue better. It was to qualify earlier, and to be honest about what a real opportunity looked like: a sponsor inside the client project team, someone at client leadership who wanted it to happen, a consultant willing to engage with the technology, and an integrator prepared to carry it. Where those four were not present, we did not have an opportunity. We had a demonstration — and a demonstration cost fifteen hundred dollars and a working week.

Two decisions, though, were ours alone.

We built permanent capability to fit the business we had, not the business we were creating. When the temporary arrangements finally had to become real — an office, a demonstration facility — we sized them around immediate requirements and then paid to expand and rebuild within a short period. Having applied "prove it before you commit" carefully to the market, we failed to apply the opposite principle to ourselves. Temporary capability should be lean. Permanent capability should be designed to grow.

And we let the first pipeline stand in for a view of the region. Our strongest early opportunities were in the UAE and Qatar, so when the company could add sales capacity, we placed it where the visible wins already were. It was rational, it produced more wins, and that is precisely why nobody challenged it — including me. I placed the resource where my own evidence was strongest, which is not analysis. It is preference wearing analysis as a costume.

Saudi Arabia continued to be covered by visits rather than by someone accountable for it. The market did not begin to ramp until we appointed someone in-country in year three — and it went on to become larger than the UAE business. The cost was not only the delay to the appointment: the relationships and pipeline still needed time to develop after the person arrived. The ramp does not begin until the resource does, so a three-year delay in the decision is a four- or five-year delay in the revenue.

What the Third Leg Proved

Within roughly a year, the Middle East was contributing to global revenue at approximately the same level as the Americas and Europe — and it held near that level for much of the period through 2015.

That is the number worth keeping, and the comparison is the whole of it. The other two legs had a head office, an established team, familiar procurement and a home-market advantage. The third leg had a rented business-centre office, a screen that travelled in boxes, and one person until September.

The percentage is not the transferable lesson. The transferable lesson is that the gap between those legs was closed by sequence, not by budget.

A greenfield market is not opened by choosing a distributor, scheduling visits, or producing a regional forecast. It is opened by understanding how a purchase actually happens there — and then positioning yourself at the point in that process where the decision is genuinely made, which is almost never the point at which the tender appears.

We built the Middle East operation in stages: visibility before product, relationships before infrastructure, technical proof before scale, and local capability at the point where the work justified it. Every investment followed evidence.

Where This Started

The work that Connektions MEA now does began in March 2010, in a rented office in Dubai, as the answer to a question a Californian start-up could not otherwise solve: the board had approved a third commercial leg in a region where the company had no entity, no employee and no way to create either.

The company was formed fifteen years later. The method did not change.

Readiness before exposure. Evidence before permanence.