Thursday, October 16, 2014

What bird droppings teach you about business

I have started posting my blogs on linkedin as an experiment.


The country of Nauru is, literally, built on birdshit. For thousands of years, this tiny Pacific island hosted millions of birds whose droppings became a mountain. In the 20th century this became a goldmine for the 10,000 inhabitants, who realized it could be extracted for fertilizer. In the 1970s, tiny Nauru was one of the richest countries in the world. Their solitary ringroad became a racetrack for their sports cars, and their investment fund became the go-to source of funds for Hollywood films when no one else was interested.
Then the birdshit ran out...
The rest of the blog is here: 

Wednesday, June 4, 2014

Online marketing is all about performance

There are two simple messages from the history of the online industry since 2000:

1. The Internet is free and funded by advertising; 
2. Advertising is about post-click performance.

While some people consider the second point controversial, the evidence is now compelling. The two behemoths of the industry, Google and Facebook, rest on this business model.

Google has grown to $40bn+ revenue per year, by selling a product "search" that delivers clicks to your site, which then convert into sales. The pitch has always been very simple: spend more money, make more money.

Facebook has experimented with many different advertising models, but has seen its revenue (and stock price) explode with its mobile news feed advertising product. Mobile is now more than half its business. The model is again post-click and performance: the advertiser pays for clicks, and measures the cost of a new app install. This focus on the cost of an app install reflects the lifetime value that is generated from installing this app (if you don't believe in this value take a look at king.com's IPO prospectus and the $1.9Bn of revenue they generated in 2013. )

At Criteo, post-click performance has been our guiding light since the very beginning. Five years ago, people used to tell me that you couldn't make display advertising perform like search. I don't hear that anymore. But performance is an ambition not a destination - it can always be improved with better technology, more data and better integrations. And the good news is that it creates great alignment with all parties involved: increasing performance for advertisers allows them to spend more money with you profitably.

I'm therefore particularly pleased with today's announcement of a dramatic improvement to the Criteo Engine - the technology that predicts and recommends what ads to show, to which users, and at what price. By being able to predict not just clicks, but conversions, we have delivered a 38% increase in the number of sales our clients can generate at the same cost of sales. Huge credit to the Engine R&D and product team for all their work on this - and it has been a lot of work! A lot of the fun at Criteo is geeky conversations about how to improve performance based on new features or capabilities, and it's nice to be able to be public about a big improvement that has come about from this.

Thursday, May 1, 2014

Criteo releases full personalized mobile advertising solution


We made public yesterday the general release of our full personalized mobile advertising solution.

This is the result of a lot of work by the product and R&D team over the last year, and it's amazing to see the difference from 12 months ago, when much of this was powerpoint!
Since we are all about results, I was particularly pleased that we could announce that Criteo is now generating over $1 billion in post-click mobile sales for its customers on an annual run-rate.
We got some nice coverage, including this in-depth piece from James Robinson at Pando

Sunday, April 20, 2014

What did the Romans ever do for us?

I visited the Roman Baths in Bath this weekend. It is a beautiful place, but this time I was struck by the fact that the reservoir and water management system that maintains the baths is the original Roman engineering. Water has been flowing for 2,000 years: through the dark ages, the age of chivalry, the industrial revolution and the atomic age, and still we were walking around on top of that Roman engineering, and watching the over-flow water pour through the great Romain sluice


Today in an age when we build for obsolescence in just a few years, we would call this massive over-engineering. Our financial experts tell us to discount our investments heavily for each additional year into the future: as a result the value of anything in 50 years time is almost zero.
Our ancestors thought differently. Bath was a pilgrimage site apparently, but it was still a tiny town on the very edge of the vast Roman empire. What I visited was Roman engineering for Peoria, Illinois. The only conclusion is that everywhere, Rome built with an extraordinary quality of engineering. The beautiful three layered viaduct of the Pont du gard is another example of Roman engineering for a run-of-the-mill urban centre of 60,000 people (Nimes) that I remember visiting many years ago:
Source: http://upload.wikimedia.org/wikipedia/commons/2/23/Pont_du_gard.jpg

Perhaps our ancestors have something to teach us. Infrastructure doesn’t survive 2,000 years by chance. The Romans must have been building for the ages. I don’t know how much written evidence we have for the philosophical world view of these lowly Roman engineers at the fringes of the Pax Romana, but it seems evident that they valued the quality of their engineering beyond its purely practical results. Would anyone have noticed if those engineers had cut a few corners and the whole edifice started to crumble 100 years later? I can only assume they built it right, because they believed it was the right thing to do. To stretch it further:  imagine arriving in this remote backwater of Northern Europe, with the knowledge and tools to transform the natural environment around you. There must have been something extraordinarily exciting to be those engineers building these great structures that would last long after they themselves were gone.

So does this mean anything for us? I think it does. I traveled to Bath on a railway line built 150 years ago by the great Victorian engineer Isambard Kingdom Brunel. Trains still race through his great tunnels and over his bridges. In fact in the West, many of us live in cities where we rely heavily on infrastructure built in the Victorian period. I live in a London house that like millions of others is over 100 years old, that uses sewers of the same age, and I travel on many underground lines that were already finished in the early 20th century. Inhabitants of New York, Boston or Paris have the same experience. Our Victorian forbears also built to last, and the great cities of the world all benefit from this approach.

So what has changed? We are far richer than the Romans or the Victorians, and yet our ambitions in the Anglo-Saxon world seem to have shriveled. Today, it is viewed as madness to build infrastructure with a long-term view - despite the fact that borrowing is at historically low rates of 1 or 2% per year. There is enormous political resistance in the UK to building the first high speed railway line to the North, despite the fact that our grandchildren and their grandchildren will undoubtedly be benefiting from it. In the US, spending on infrastructure has fallen dramatically and most visitors to New York City are struck by the appallingly poor transportation infrastructure for such a great metropolis, where it is impossible to get from JFK to Manhattan by a fast direct train, and where the subway leaves big swathes of the island poorly served.

The main streets around me in London are dug up 2 to 3 times a year, bits of infrastructure (water, gas, electricity, cable…) placed in the clay and then covered over. It’s a process the Romans would be entirely familiar with. It seems hard to believe it wouldn’t make more sense to spend more upfront, build proper ducting for these services and those that will come (fibre to the home?) and never need to dig up these roads again.

Closer to my own working life: in the Internet we have some of these same issues. Too often in the tech industry we worship businesses for the price for which they are acquired, rather than the long-term impact that they have. The Valley talks a lot about culture, but in fact all that dynamism creates remarkably few businesses that last, and even fewer with a strong culture. It seems inevitable that as a result we tend towards an engineering culture that does not build for the long-term, and focuses on “good enough” instead of excellent. Apple is a great counter example, with both a very strong culture and the powerful idea that its products must be excellent before they are released – I suspect that for this reason Apple after Jobs will continue to be much more potent than its detractors claim.

I believe much of this should be blamed on the culture of finance, and of shareholder value measured by the current stock price, that became so popular starting in the 80s. If the “truth” is that everything should be discounted back to net present value at 10-15% per year, then of course the future doesn’t matter very much. The fact that our children will live in this future doesn’t really figure. Similarly too much focus on the near-term stockprice often means we reduce longterm value creation: I truly believe that the best way to create value for shareholders is to focus on creating longterm value for customers.


Of course this message is overly simplistic. London is currently building the massive Crossrail project (very cool website if you like your engineering physical not just digital - check it out), that will undoubtedly provide infrastructure we are still using in 100 years. And outside the anglo-saxon sphere, France is famous for the network of high speed rail it built in the 80s, and German cities are always noticeable for their efficient airports and network of autobahns. But despite this, I feel there is something to learn from those Roman baths and the engineers who just wanted to build something right.

Monday, March 31, 2014

2014 is the year we'll fix mobile marketing

Adweek has landed in Europe, and as well as joining lots of panels this week, I've written a little piece for them. For consumers the "year of mobile" was obviously a couple of years ago, but interestingly mobile marketing is fairly broken for most advertisers. I've tried to explain why I think 2014 is the year we'll fix mobile marketing: Mobile App Metrics: Why 2014 is the Year of Mobile Marketing

Lots of thanks to Jason and Mike in my team who have done much of the real thinking behind this.

Thursday, January 23, 2014

Releasing Performance Display Advertising for mobile apps

Today we released our Performance Display advertising solution for mobile apps. My fantastic Product and R&D team at Criteo has been working on this for almost a year now, so I'm really excited that we have been able to deliver this and go public about it. I'm also delighted to see the impact of the team from Ad-X Tracking, which was instrumental in delivering a fully functioning solution.

There is a great article at the FT providing a lot of context, but you will need to register. It is also covered at Adweek and elsewhere. (Update: also at AdExchanger)

I'm particularly pleased because a lot of people said to me that the sort of individually personalized performance advertising that Criteo historically delivered on the desktop couldn't work on mobile. I guess I never like being told something isn't possible, and it seems neither does anyone I work with...

I think this whole area of mobile commerce and advertising is particularly interesting, because we are seeing a big platform shift right now. In many ways this is the end of a long period of stability - it is 20 years ago now that Netscape was founded, and the Internet since then has been about a browser sitting on top of a Windows-Intel dominated PC. All the great e-commerce and website/advertising businesses that we think of were built on this combination of browser and desktop: Amazon, Ebay, Google, Yahoo and even Facebook. It is less than 7 years since the first iphone was launched, and only a couple of years since smartphones became broadly adopted but we have moved to a new era, where mobile is very rapidly becoming the dominant platform. And on this platform, while the browser remains important, the majority of a user's time is being spent in mobile apps. As a result there is a great deal of disruption going on.

We all know there are a huge number of users on smartphones now, but the most dramatic shift has been the recent growth in transactions on these devices, and the advertising that follows from this. I discussed this earlier this month when looking back at 2013's megatrends, and in particular the fact that 25% of all US Thanksgiving sales were on mobile devices, and half of Facebook's revenue was also on mobile devices.

Until now the mobile advertising business has been almost entirely focused on app instals, and heavily focused on games. Inevitably, the cost of new user acquisition has been rising and the focus starting to shift towards ensuring that all this money spent on acquiring customers ends up being profitable. That means ensuring that users continue to engage with the app after it's installed, and figuring out how to increase conversions.

Monday, January 13, 2014

My megatrends of 2013

Note: These trends are written in a personal capacity, and are based on publicly available information.

With 2013 over, I thought it would be interesting to list the 5 biggest trends in the ecommerce and online advertising industry. Everyone is likely to have a slightly different list, but here is mine:

1. Mobile commerce happened while you weren't looking. Do you remember all those articles in past years explaining that Japan was unique because they used mobile feature-phones to buy things? This is the year everyone became like Japan. Mobile sales were over 25% of all sales in the US during thanksgiving according to IBM

2. Mobile advertising works just fine. Facebook went from being pilloried for no mobile strategy to being lauded for doing half its revenue on mobile. Of course total global mobile advertising revenue is still small, but 2013 proved that most of the technical issues are now solved and that this is now mainly about adoption.

3. The Online “Pivot to Asia” is dramatic. The global centre of gravity for online commerce and therefore advertising has been steadily shifting to Asia. For a long time Japan, Korea and Australia have been highly penetrated online markets. However the growth in emerging markets and particularly China continues to astound. This was made concrete in 2013 by Alibaba’s announcement that on one day its sales reached almost $6Bn
This is driven by two trends that have continued strongly in 2013. Since 2008 GDP in developed countries has been basically flat due to the crisis; in China it has increased 50%. This graph from the Economist says it all:



Alongside this the second trend is the stupendous switch online in China . In 2013, China had 38 million new online shoppers, more than the population of Canada, according to BCG . As a result the total Chinese ecommerce market was $145Bn  in 2013.

4. Android wins big. In 2012 we were all told only Apple monetized for developers, and so Apple would be the winner of the apps war, and therefore the mobile war. Today we read articles claiming that Google’s Android may be to mobile what Microsoft Windows was to the PC. The monetization gap still exists but it was closing fast in 2013: Google’s app revenue was only 42% of Apple in June, but had grown to 59% by November . Meanwhile the scale of new Android devices in 2013 continued to dwarf everything else, three times as many as Apple shipped. This isn't to say 2013 showed Google winning this race - but it definitely became a race. Those people who said Google was a one-trick pony (with a good trick), and that big companies can’t innovate, need to look closely at this. In fact Google’s overall performance in 2013 was tremendous.

5. Digital continues to eat analog advertising. Finally, the trend that’s so obvious we tend not to discuss it. In developed markets where GDP has been almost flat, online advertising has continued to grow at a tremendous rate. In the US alone, it grew by 15% Year on Year. While online at 22% of US spend is still below TV, it continues to grow much faster. The growth of online advertising continues to be a megatrend that has driven a lot of the innovation of 2013, and funds the web giants outside of Amazon and Apple. 


Those are my 5 megatrends of 2013. Two trends almost made the list: video, and Google's performance. On video, I would argue that 2013 was actually a bit disappointing when you look at the size of the TV market, and the fact that 88% of digital video advertising was on Youtube. Google's online performance almost made it though, with both Android and Youtube making huge strides in 2013. According to Goldman Sachs, 2% of all advertising moves online each year and Google will take half of this. 

My overall conclusion is simple: we are seeing a very high level of change online compared to 5 years ago (and we thought things were moving fast then!). How different will my trends for 2014 be?