On 23 April we held our first Rocket Fuel summit.  Hopefully, after taking on some of the eponymous fuel, our attendees will feel ready  to fly ahead with their data strategy. For our part, it was a tremendous success. We're extremely proud of producing such an exciting event.

We at Nimbus Ninety would like to extend an enormous thank you to everybody who came, who made it so exciting, so engaging, and such a joy to host. In particular, we’d like to extend a massive thanks to our speakers, our partners, and our panel.

We wanted to build a highly-distilled summit; hyper-concentrated. It was a double espresso, not an americano. It was jam-packed. Therefore, below, we have done our best to compile some of what was said, but we can only offer a taste.

What We Learned

Storage is cheap, data is scarce; so store everything. You never know when you might need it.

STEAM is the new STEM. The A is for “Arts.” Maths is not the only discipline important for your business. As customers become the centre of businesses, creating products that customers like the look and feel of is more important than ever. Besides, AI can do maths; arts graduates may be more and more important.

Neural networks can yield black box results. This is a problem. There will be a lot of political issues with AI making this or that decision without explaining how or why. We need to build models which enable AI to explain what decisions they make; so that you can explain them to customers.

The really exciting AI work at the minute is around pictures and videos. With the possible exception of voice, pictures are the most ubiquitous form of “unstructured” data we have, which AIs can now parse. There are so many exciting breakthroughs coming out in this space. Speaking of which,

You can get life insurance quotes from selfies now. Do you know how much data is locked up in a selfie? Age, gender, lifestyle, and more. Expect more developments like this in this space.

Data-led is better than data-first. Data is pointless if you don’t react the right way. You need to actively work out to what B you want to get from the A at which you currently stand.

Document, document, document. Within the changing digital policy landscape, documentation is the easiest rule of thumb to minimise risk exposure.

The Clergy probably won't be replaced by robots. Although there is a robot priest operating in Germany...

Test More Variables. A/B? Forget it. What about A/B/C/D/E. Did you know that several variables will make the employees feel like empowered decision-makers rather than conduits in a flowchart? Did you know this will enable them to test riskier ideas which they really think could work?

There is lots of paper-led backend in tech-phobic businesses. These are ripe for disruption. The CEO of Habito explained how antiquated much of the mortgage industry is. Even introducing great UX and working APIs can change some industries.

Tech-phobic businesses generally won’t change unless they’re forced. If you want to bring change, a struggling business is more likely to listen.

Most businesses in the UK today are “tech supported”. The leap to “tech-driven” is perhaps the biggest and valuable a company can undertake.

GDPR is an opportunity as well as a risk. It has forced businesses to think about the value of the data. The companies that will do the best out of GDPR will use it as an excuse to stop collecting bad data and re-engage your customers.

Data is not oil; it’s water. The more you have; the more you can explore, and the more “clams” of insight you can find down there.

Data collection can be framed in terms of exchange to consumers. People understand their data is valuable; but like their money, they will be willing to part with it, if you have a great service and a great reason for doing it.

There are lots of blockchain sceptics. Will blockchain dating change the world? Rory Cellan Jones is sceptical. Only slightly more sceptical that he is about the significance of a currency that can change it’s value by 15% in 40 minutes.

Our Favourite Three Takeaways

1. You cannot expect iterative returns on data. It’s exponential.

Some businesses expect that if they invest a little money in data, they’ll see a little return. This is false. But the attitude can disincentivise AI investment. Unlike with software, you must commit to ML/AI for a long time before it adds value. For this reason, many projects are dropped before explosive results are realised. A serious approach to data will see serious returns.

Before you realise these, you must untangle and clean your data, or as one speaker put it, “unf*ck” your data. For many uses, you must have perfect data to function. Who would want a bank statement which is correct 98% of the time?

2. GDPR: Global, Noble, and Unknown

Some companies have migrated their clouds to the US to escape European Data regulation. This could be short-termist. Our experts pointed out that America may implement very similar policies in the coming years.

GDPR was characterised as championing a noble cause, but being already unfit for purpose. This isn't necessarily the fault of the regulators; law will always be playing catch-up with tech as ‘you have to wait for it to be invented first!’

The UK ICO is already gearing to use its powers; but how much warning and hand-holding they will give around the law remains to be seen. The most likely scenario is some kind of warning and preventative actions with an additional compliance period before a fine is levied.

3. Business is about timing and customers

Before they lost the phone market in its entirety to Apple, Nokia released a kind of multi-function phone. But nobody bought it. Being early (or late) is the same as being wrong. And being innovative is not always as important as getting the fundamentals of the product right.

A big part of the model of incubators is telling people not to get distracted, not to think about what they want, or their own ideas, or things from the perspective of the business. It’s about asking again and again and again, what do your customers want? Is this what they want? How can we make it better? Data, for all the hype, is just a tool to aid this. This is the goal. This is the main thing. One CEO had everyone in his business do customer service shifts. There is no replacement for knowing your customer and what they want, now.