Nokia saw the iPhone coming, but couldn’t act on what it knew about itself

On 10 January 2007, the morning after Steve Jobs held up the first iPhone in San Francisco, people inside Nokia met to work out what they had seen.

The slides from that meeting are public now, held in the Nokia Design Archive at Aalto University, which opened on 15 January 2025. The executive summary calls the iPhone a serious high-end contender. Further down, the deck records that the user interface had been a strength for Nokia, that consumer research showed it slipping, and that Nokia needed to develop touch in order to fight back.

Nokia didn’t fail because it couldn’t see what was coming. It saw the fault line the next morning.

It had seen it a year before that.

In 2006 the computer scientist Jeff Han gave a TED talk demonstrating a multitouch screen. Timo Partanen, then Nokia’s director of market and competitor analysis, burst into a colleague’s office telling him he had to watch it. The colleague was Peter Bryer, Nokia’s manager of strategic foresight. Both of them thought it was the future. Then Bryer looked at who had paid for Han’s research, and Nokia was one of the sponsors.

Nokia’s first multitouch phone shipped in 2010. Four years after the talk it had helped fund.

Bryer’s explanation for the delay is the most useful thing anybody has said about this story. Nokia is in Finland. It is cold in Finland. People wear gloves for six months of the year, including the executives, and they did not believe a device like that would work.

Set that against Nokia’s own accounts. The ten markets where Nokia sold most in 2007 were China, India, Germany, the UK, the US, Russia, Spain, Italy, Indonesia and Brazil, and between them they made up around half of net sales. The people deciding whether a touchscreen was viable were testing the question on their own hands, in Espoo, in winter.

Nobody in that room was foolish. They were the sample they had.

Partanen was at the meeting the morning after the launch too, and his recollection is that there was little concern in the room. Another competitor had launched a good product. If it worked, Nokia would do the same.

The numbers agreed with him.

Ten months later, in the first week of November 2007, Nokia’s New York-listed shares reached a high they have never returned to. The company accounted for around a third of the entire market capitalisation of the Helsinki exchange. Gartner would report that Nokia took 40 per cent of the global handset market in the fourth quarter of 2007, the first time it had crossed that line, after selling 435 million devices across the year. Interbrand and BusinessWeek had already ranked Nokia the fifth most valuable brand in the world at $33.7 billion, the highest-placed non-American name on the list.

Nokia’s own results for the year showed net sales of €51.1 billion and operating profit of €8.0 billion. The three device business groups produced €7.9 billion of that between them, and Mobile Phones alone ran at an operating margin of 21.7 per cent. In the fourth quarter Nokia shipped a record 133.5 million devices. Olli-Pekka Kallasvuo, the chief executive, called the quarter excellent.

Every one of those numbers was true. Every one of them was measuring the canopy.

The distance between what an organisation knows somewhere inside itself and what it can accept, decide and act on has no line in the accounts.

Nokia’s own people had correctly identified the fault line in January, and the warning didn’t disappear. Neither did it become action proportionate to the threat. It was absorbed into the existing plan: develop touch, keep S60 at the centre, trust Nokia’s scale and engineering capacity to close the gap.

The company understood what Apple had built. It had not yet worked out what Nokia itself had become.

The distance between what an organisation knows somewhere inside itself and what it can accept, decide and act on has no line in the accounts. It carries no recognised liability and attracts no auditor’s note. Yet it can decide the future of the whole enterprise while every reported measure is still improving.

Very few readers of this run a business holding 40 per cent of a global market. The mechanism doesn’t care about size, and it works the same way in a twelve-person engineering company. Somebody sees something important. The information is softened as it travels, interpreted through what leadership already believes, or acknowledged without anybody accepting the cost of acting on it. By the time a decision is made, the organisation is responding to a safer version of the truth.

The most careful account of what happened inside Nokia comes from academic research rather than contemporary reporting. Timo Vuori of Aalto University and Quy Huy of INSEAD interviewed 76 people, including senior executives, middle managers, engineers and outside experts. Their study, published in Administrative Science Quarterly in 2016, examined the period from 2005 to 2010. It begins two years before the applause peaked.

What they describe is shared fear moving in two directions. Senior managers feared competitors and shareholders, and that fear translated into pressure on the people beneath them without a full explanation of how serious the external threat had become. Middle managers feared their superiors and their peers. What they feared was less the loss of their employment than the loss of standing: appearing weak, obstructive, disloyal or unable to deliver.

So the bad news softened on its way up.

Engineers who understood the scale of Symbian’s limitations communicated them in ways that protected their position. Deadlines that people privately regarded as unachievable stayed in the plans. Comparisons between what Nokia’s software could actually deliver and what senior management believed it could deliver became progressively less reliable. The result at the top was a picture of Nokia’s technological capacity that was better than the capacity itself, and capital, deadlines and product commitments were allocated to match the picture.

What was decaying was not the product line and it was not the balance sheet. It was the set of conditions in which information moves: whether somebody can say a hard thing to the person above them and still be in the room the following week. Nothing in the accounts records the state of those conditions. They were holding up everything the accounts did record.

One senior manager told the researchers he could not say publicly that Symbian was finished, because of what that admission would do to Symbian sales.

“You must believe in the gun you are holding,” he said, “because there is nothing else.”

That is the economic trap at the centre of the story. Nokia had to hold two contradictory truths at once. Symbian was still producing billions of euros in revenue and profit, and Symbian was also becoming incapable of securing the company’s future. The organisation could sell the first truth or act on the second. It could not easily do both.

Calling the platform structurally compromised would have weakened confidence among customers, operators, developers and investors while Nokia still depended on it. It would have challenged years of investment, the standing of senior leaders and the influence of a large internal organisation built around Symbian. It would have required Nokia to damage the current profit engine before a replacement was ready. The question was therefore larger than whether somebody felt safe enough to speak, because the truth itself carried a price.

None of that appeared in a quarterly result for a long time. Nokia’s handset volumes kept climbing into 2008. The market share, the margins and the brand remained formidable. The reckoning ran on a different clock.

Nokia demonstrated a touch interface for S60 on concept devices in October 2007. Its first touchscreen S60 handset, the 5800 XpressMusic, arrived in November 2008 with a resistive screen and no multitouch. Partanen’s account is that it was delayed and delayed and delayed and arrived watered down. It sold around eight million units in its first year, which by any ordinary standard is a success, and it did not change anything.

The company had not ignored the signal. It had translated a change in the basis of competition into a feature-development programme for the platform it already owned. Apple had made software, interface and the relationship between hardware and software central to the value of a phone. Nokia responded as though touch were an additional capability that could be fitted onto its existing advantage.

The distinction consumed the time.

In the first quarter of 2012, Samsung passed Nokia to become the world’s largest handset manufacturer, shipping 93.5 million devices against Nokia’s 82.7 million. Nokia’s run at number one, which had lasted since 1998, was over. Standard & Poor’s and Fitch both cut the company’s debt to junk status in the same week.

In September 2013 Nokia agreed to sell substantially all of its Devices and Services business to Microsoft. Microsoft paid €3.79 billion for the operating business and a further €1.65 billion for a ten-year licence to Nokia’s patents, €5.44 billion in total. The transaction completed in April 2014. In July 2015 Microsoft wrote off $7.6 billion against it, close to what it had paid.

Put two of those numbers beside each other. Nokia’s device businesses generated €7.9 billion of operating profit in 2007. Six years later the whole operating business, plus the patent licence, went for €5.44 billion.

The visible collapse arrived quickly. The conditions behind it had been developing while the visible performance was still exceptional.

Which leaves the question this piece exists to answer. What could an outsider have seen at the time, from the public record, dated on or before Nokia’s peak?

Not the academic study, whose first interviews were years away. Not the January presentation, which stayed inside Nokia. Not the internal deadlines, the softened reports or the private assessments of Symbian’s condition.

The honest answer is very little.

Three things were visible. On 20 June 2007, Nokia announced that it was taking apart the structure of its devices business and rebuilding it as three units from 1 January 2008. Its spokeswoman, Arja Suominen, said it was not a cost-cutting exercise and that the company was seeking greater efficiency. It was reported largely as organisational housekeeping.

On 16 October 2007, at the Symbian Smartphone Show in London, Nokia demonstrated a touch interface for S60 running on concept devices. Anyone with an internet connection could watch the video. Nine months had passed since the iPhone was unveiled and four since it had gone on sale, and the world’s largest phone manufacturer was showing a prototype response.

The third was the one nobody joined up. Jeff Han’s multitouch demonstration was a public TED talk, and Nokia was listed among the sponsors of the research behind it. By November 2007 an outsider could have known that Nokia had helped fund the interface technology at the centre of Apple’s product, and that Nokia had so far shown a resistive prototype without it. That is a real signal and it was freely available.

It was also nowhere near strong enough. The public could see that Nokia was reorganising and developing touch. It could not see that deadlines were regarded internally as unrealistic, that information was being softened on the way up, or that leaders held an inflated view of Symbian’s readiness. The prototype was received as evidence that Nokia was responding, rather than as evidence of how far behind it might already be.

There were clues. There was no public signal powerful enough to outweigh record volumes, market share, margins, profit and brand value. Almost everything conventional analysis measured in November 2007 pointed the same way.

There was no visible crisis to detect. The crisis was in the system through which reality became a decision.

For investors, acquirers and incoming chairs reading this, that is where it matters. Financial diligence can tell you what an organisation has produced. It is far less capable of telling you whether the people closest to the work can contradict the assumptions behind the forecast, whether the contradiction reaches the people allocating capital, and whether those people are willing to bear the cost of acting on it. Does the business have a working route for bad news to reach the person who can act on it? Does the information survive the journey intact? Can leadership tell a manageable product problem from a shift in the basis of competition? Has anybody tested the route using information that threatens current revenue, standing or sunk investment?

Samsung met the same disruption, in the same market, at roughly the same moment. It is not a controlled experiment. Samsung had different assets, less software legacy to protect, substantial display and semiconductor capabilities, and fewer reasons to defend a single proprietary smartphone platform. That difference is part of the point.

Across 2007 Samsung was the third-largest handset manufacturer, shipping devices running Windows Mobile and Symbian. It announced its own operating system, Bada, in November 2009 and carried on developing it. It also put its flagship smartphone on Android. Samsung launched the Galaxy S in June 2010 across roughly 110 operators in 100 countries, accepting that its most important handset would run software controlled by another company, and that the success of Android could make its own platform irrelevant.

That was the cost Nokia’s managers feared. It meant surrendering a degree of control, weakening the case for proprietary software and redirecting investment away from capabilities the company had spent years building.

Samsung paid the price of depending on Android. Nokia went on paying the price of protecting Symbian.

Nokia’s past success made the necessary decision more expensive. The larger the revenue stream and the internal organisation and the accumulated identity attached to the existing model, the harder it becomes to treat evidence against that model as an instruction rather than a problem to be managed.

The transferable lesson has little to do with mobile phones or platform strategy. The quality of a business’s decisions is capped by the quality of the unwelcome information that reaches its leaders, and by their willingness to bear what acting on it will cost.

Where honesty carries a personal charge, bad news will soften. The charge need not involve dismissal or formal punishment. A cooler tone, a shorter hearing, an irritated response, exclusion from the next discussion. Any of those is enough to teach somebody what kind of truth is welcome.

Where the truth threatens current revenue, internal standing or sunk investment, its implications can soften even when the words arrive intact. A leader can hear the warning, thank the messenger and carry on funding the same plan. Creating safety for the messenger costs very little. Acting on the message may cost a great deal. It may mean abandoning investment, cannibalising a profitable product, changing a public commitment, or admitting that the capability the whole strategy rests on is weaker than leadership believed.

That is the test, and it can be checked this week.

Bring to mind the last three occasions when somebody reporting to you brought unwelcome news: a date that had slipped, a product that wasn’t working, a competitor pulling ahead, a customer preparing to leave, a capability weaker than the plan required. Did the messenger pay for it, through your tone, the length of the hearing, or the access they got afterwards?

Then ask what changed because they told you. Which decision was reopened. Which resource moved. Which assumption was tested. Which commitment was stopped.

If the answer is that you listened carefully and carried on, the information arrived and the truth still didn’t. And if you can’t bring three occasions to mind at all, the absence is the finding.

Then look at a number you already watch. Nokia’s leadership held a picture of its own technical capacity that was better than the capacity, and it committed capital and deadlines to the picture. That gap does not show up in revenue for years. It shows up almost immediately in delivery. Take the dates in this year’s plan and ask how many of them were set by somebody who believed them. You can’t audit that directly. You can look at the proportion you have actually hit over the last eighteen months, and at whether it has been drifting in one direction for longer than anybody has said out loud.

Somebody in your business already knows what an analysis like this would uncover. They may have explained it carefully, softened it into language the organisation can tolerate, or decided that the consequences of saying it plainly are worse than the consequences of staying quiet.

The question is whether you have taught them that saying it is safe. The harder question is whether you are willing to pay what acting on it will cost.