How HP valued Humane

Last week HP announced that they were acquiring the patents and key staff from Humane for $116 million.  A few years ago, I wrote an article questioning the value of patents for startups (of which more later), so it seemed a good opportunity to try to dissect the purchase price to see if it’s possible to put a value on Humane’s patent portfolio.

Humane’s not your average startup.  From the start it was viewed as a potential unicorn.  Its first product – the AI Pin, which turned out to be its downfall, had high ambitions.  Although few reviewers appeared to notice it, if it had succeeded, it would have been the first nail in the coffin of the smartphone.  The product failed to meet that promise, and the company appeared to be heading for a fire sale.  Fortunately for the core team, HP saw their value and snapped them up for the bargain price of $116 million.  Let’s look at how they might have worked out that purchase price.

Buying small or faltering startups with promise has been a standard strategy for larger tech companies trying to inject new life into their development teams, or jumpstart new developments, or enter new areas.  The age of multi-billion dollar acquisitions led to people forgetting that there used to be some basic metrics to working out a value, which are:

• How many key visionaries are there that you can persuade to stay with you for at least three years?
• How many of the core engineers need to come with them?
• What’s the value of their customer base?
• What’s the value of their ongoing business, assuming it’s useful?  And
• What’s the value of their IP?

In the late 1990s and early 2000s, when semiconductor expertise was thin on the ground, the going price for key team members in such acquisitions was around $1million per engineer.  As we’ve gone through various tech hype cycles, that price has risen.  Big Data took the key engineer value to around $3million by 2010.  With the current frenzy for AI, I suspect it’s doubled – up to $6 million per head.  At that level, there won’t be many candidates in any company – it will be the key visionaries and leaders.  I’m guessing there will be around five for Humane, accounting for the first $30 million of the acquisition price.

The value of picking up the expertise of a company like Humane comes in keeping the core developers as well, as that gives you many man-years of high class development work, which is almost impossible to generate from a standing start.  If you’re trying to integrate a new group into your organisation, you’d be looking to retain about 20 key people, valuing them at around $1.5m each.  You normally don’t want more, as it’s easier to integrate a smaller team.  Bear in mind that this isn’t their salaries, but the value a company would put on the expertise they’re acquiring when they try to justify the budget for their purchase.

Adding those together, for that $60 million, HP will get a highly functional AI team, whose development platform is at the bleeding edge.  There aren’t many of those up for sale, and in today’s frenzied AI world, that feels like a bargain price.  HP will need to incentivise them to stay as a team, but that sounds as if it’s already happening, with significant salary rises to keep the core team alongside the key players, who presumably already have attractive tie-in packages.

In Humane’s case, I doubt HP sees any value in their current customers, or are planning an AI Pin v2.  So neither the customer base nor the product have any value.

In all of these deals, the winners tend to be the lawyers and advisors that put it together, who typically take an eye watering amount of money.  For something of this scale, it’s probably around 26% of the deal, which will net them about $24 million.  Once we factor that in, we have a nominal value for the patent portfolio.  The deal was $116m, we’ve accounted for $84m of it, which leaves a value of $32 million for the patent portfolio – a nice round $100,000 for each of the estimated 320 patents.  That probably isn’t much more than each of them cost, once you’ve factored in the engineering time and patent attorney fees.  Had that time and money been spent on making the AI Pin better at launch, it might all have been a different story.  But given Humane’s burn rate, it might not.

For Humane’s investors, it’s not a great outcome, given they’d invested almost a quarter of a billion.  But VC investment at this scale is largely roulette with better opportunities for tax deductions.  Rien ne va plus.

My original article on whether startups should spend money on patents resulted in more comments than anything else I’ve written.  What was interesting is that most came as private emails, rather than being added publicly to the article.  The general feeling is that startups should concentrate on delivery rather than IP.  The main exception to that view was from CTOs, who mostly felt that a granted patent was a global acknowledgement of their genius, which a grateful company should take all steps to procure.

I wasn’t considering proto-unicorns like Humane when I wrote that article.  Companies like Humane are funded to burn brightly, either making it to the elevated plains of unicorndom or crashing with their investors’ hopes and money.   However, looking at the acquisition value, it feels that the same law of patent value versus delivery probably applies to proto-unicorns as smaller companies.

At the end of the day, it’s sad that the AI Pin didn’t manage to accelerate the final stages of the life cycle of the smartphone, but it’s not a bad exit for the core team.  I know from personal experience that being acquired by a large corporation just before a market bubble bursts can be excellent timing.  It won’t buy you a desert island, but it gives you stability to reassess and evolve your core technology while your competitors are forced to focus on day-to-day survival. 

For HP, they’ve got a ready-made team that’s at the top of their game.  If they use it well, it should more than pay for the acquisition.  Whether the patent portfolio is useful, we may never know.  But at least it’s in responsible hands, rather than having been picked up as tasty treats for patent trolls.  Sadly, that’s the fate of many startup patents, which then come back to spread anguish and pain amongst the next generation of startups.

All of this, of course, is pure speculation.  HP may have acquired Humane for any number of reasons.  But since their experience with Autonomy, I’d hope they’ve learnt a bit more about due diligence, so I suspect my numbers have some validity.  I wish both parties well and look forward to seeing what comes out of it.

Last week HP announced that they were acquiring the patents and key staff from Humane for $116 million.  A few years ago, I wrote an article questioning the value of patents for startups (of which more later), so it seemed a good opportunity to try to dissect the purchase price to see if it’s possible to put a value on Humane’s patent portfolio.

Humane’s not your average startup.  From the start it was viewed as a potential unicorn.  Its first product – the AI Pin, which turned out to be its downfall, had high ambitions.  Although few reviewers appeared to notice it, if it had succeeded, it would have been the first nail in the coffin of the smartphone.  The product failed to meet that promise, and the company appeared to be heading for a fire sale.  Fortunately for the core team, HP saw their value and snapped them up for the bargain price of $116 million.  Let’s look at how they might have worked out that purchase price.

Buying small or faltering startups with promise has been a standard strategy for larger tech companies trying to inject new life into their development teams, or jumpstart new developments, or enter new areas.  The age of multi-billion dollar acquisitions led to people forgetting that there used to be some basic metrics to working out a value, which are:

• How many key visionaries are there that you can persuade to stay with you for at least three years?
• How many of the core engineers need to come with them?
• What’s the value of their customer base?
• What’s the value of their ongoing business, assuming it’s useful?  And
• What’s the value of their IP?

In the late 1990s and early 2000s, when semiconductor expertise was thin on the ground, the going price for key team members in such acquisitions was around $1million per engineer.  As we’ve gone through various tech hype cycles, that price has risen.  Big Data took the key engineer value to around $3million by 2010.  With the current frenzy for AI, I suspect it’s doubled – up to $6 million per head.  At that level, there won’t be many candidates in any company – it will be the key visionaries and leaders.  I’m guessing there will be around five for Humane, accounting for the first $30 million of the acquisition price.

The value of picking up the expertise of a company like Humane comes in keeping the core developers as well, as that gives you many man-years of high class development work, which is almost impossible to generate from a standing start.  If you’re trying to integrate a new group into your organisation, you’d be looking to retain about 20 key people, valuing them at around $1.5m each.  You normally don’t want more, as it’s easier to integrate a smaller team.  Bear in mind that this isn’t their salaries, but the value a company would put on the expertise they’re acquiring when they try to justify the budget for their purchase.

Adding those together, for that $60 million, HP will get a highly functional AI team, whose development platform is at the bleeding edge.  There aren’t many of those up for sale, and in today’s frenzied AI world, that feels like a bargain price.  HP will need to incentivise them to stay as a team, but that sounds as if it’s already happening, with significant salary rises to keep the core team alongside the key players, who presumably already have attractive tie-in packages.

In Humane’s case, I doubt HP sees any value in their current customers, or are planning an AI Pin v2.  So neither the customer base nor the product have any value.

In all of these deals, the winners tend to be the lawyers and advisors that put it together, who typically take an eye watering amount of money.  For something of this scale, it’s probably around 26% of the deal, which will net them about $24 million.  Once we factor that in, we have a nominal value for the patent portfolio.  The deal was $116m, we’ve accounted for $84m of it, which leaves a value of $32 million for the patent portfolio – a nice round $100,000 for each of the estimated 320 patents.  That probably isn’t much more than each of them cost, once you’ve factored in the engineering time and patent attorney fees.  Had that time and money been spent on making the AI Pin better at launch, it might all have been a different story.  But given Humane’s burn rate, it might not.

For Humane’s investors, it’s not a great outcome, given they’d invested almost a quarter of a billion.  But VC investment at this scale is largely roulette with better opportunities for tax deductions.  Rien ne va plus.

My original article on whether startups should spend money on patents resulted in more comments than anything else I’ve written.  What was interesting is that most came as private emails, rather than being added publicly to the article.  The general feeling is that startups should concentrate on delivery rather than IP.  The main exception to that view was from CTOs, who mostly felt that a granted patent was a global acknowledgement of their genius, which a grateful company should take all steps to procure.

I wasn’t considering proto-unicorns like Humane when I wrote that article.  Companies like Humane are funded to burn brightly, either making it to the elevated plains of unicorndom or crashing with their investors’ hopes and money.   However, looking at the acquisition value, it feels that the same law of patent value versus delivery probably applies to proto-unicorns as smaller companies.

At the end of the day, it’s sad that the AI Pin didn’t manage to accelerate the final stages of the life cycle of the smartphone, but it’s not a bad exit for the core team.  I know from personal experience that being acquired by a large corporation just before a market bubble bursts can be excellent timing.  It won’t buy you a desert island, but it gives you stability to reassess and evolve your core technology while your competitors are forced to focus on day-to-day survival. 

For HP, they’ve got a ready-made team that’s at the top of their game.  If they use it well, it should more than pay for the acquisition.  Whether the patent portfolio is useful, we may never know.  But at least it’s in responsible hands, rather than having been picked up as tasty treats for patent trolls.  Sadly, that’s the fate of many startup patents, which then come back to spread anguish and pain amongst the next generation of startups.

All of this, of course, is pure speculation.  HP may have acquired Humane for any number of reasons.  But since their experience with Autonomy, I’d hope they’ve learnt a bit more about due diligence, so I suspect my numbers have some validity.  I wish both parties well and look forward to seeing what comes out of it.