A couple of weeks ago I was asked to present at the Microsoft Analyst Summit in Singapore on how we (at MOQdigital) promote, evolve and execute on digital transformation for our customers.
I was fortunate to be sharing the stage with some great innovative thinkers in their own leagues.
After been given the pre-presentation pep talk in the speaker room of “The room is full of analysts and they can be a tough crowd…so lot’s of energy and drama…” – not quite the words I would have used by cross country boundaries, we’re all good to go. No pressure.
Darrel, Ryan and myself were talking about what we’d done recently in the digital transformation space, facilitated by Andreas. Some great material from Darrel – where at an energy generation and distribution business, he turned the business to be more customer focused and able to deploy new features into their Azure platform each 30mins. From development through to prod. A massive turnaround from the usual 6-12 weeks!
Ryan & I shared different sides of the coin from the same innovative solution – SensaMate.
Great audience and a room full of analysts which made for some great Q&A.
For me the real value came when I caught up with several of the analysts at an evening function, we’re able to kick tyres and validate our thinkings.
As we know the business world has a huge focus on analytics, collecting, cleansing and making heads or tails of it. In my opinion this is 2016!
If you’re not measuring it, then you’ve already lost the race.
The need for the Data Scientist….
I’m currently working with a couple of clients around AgriTech (another hot topic today) – and an interesting scenario hit me – We actually don’t need a data scientist as such (not full time anyway). Let me explain….
Let’s say we know nothing about farming, crops look good, they look bad, not really sure when they should be watered or not.
Using sensors in the soil to measure moisture (this can be done with the most basic Dick Smith funway into electronics kit – for those of you whom remember those :)), we measure, record and feed back the sensor data to our data set.
When the farmer comes along and waters the crop, we have data before and after watering. Importantly we have recorded what success looks like. Several months of this and we have a decent data set built up around crop hydration.
We can take this data set & solution to the farmer down the road and say “We’ve got an intelligent watering system that will save you $ per year….”.
By recording success, the data set is telling/showing us what to do, when.
…remember we know nothing about farming 🙂
What about the data scientist…. yes & no…
There are many market places opening up such as Azure Machine Learning where algorithms can be exposed as a SaaS model – we plug algorithm XYZ in, and call it as a webservice for e.g. Check the results, change to a different algorithm to suite. Very much a trial and error approach.
Data Scientists most definitely can come up with the original algorithm, model etc – but most of these are consumable with parameters to tweak them when needed.
Compare to Excel – everyone’s favourite desktop tool. We don’t go and write all the functions you use in cells from first principles, we use a library and adjust parameters to suit.
Algorithms are no different – we have the compute power, we have the marketplace where analysts can publish their genius – the fringe case where something doesn’t quite fit, then yes absolutely get the data scientist in to complete the piece.
But I see them more completing 5-10% of the masterpiece by ‘hand’ while the rest can be commoditised.
In any event, focus on getting into the position where you can use a data scientist – creating master data services, data store etc.
The beauty of the cloud is – test & learn. Build something, try it, test it, use it, take the learnings….and….it costs you $ and not $$$ as in years gone by.
The Analysts’ Top 3 concerns around Customers Digital Transformation Journey:
- Security – what level of risk/comfort are companies happy with embarking down the transformation journey?
- Digital examines many areas and levels within the organisation. This can be from simply the wish to provide BYOD through to providing 360 degree views of the customer experience.
- How agile, what security measures are in place? *Assume everything is unsecure* – just because the company’s internal network is being used, doesn’t mean it’s authorised traffic.
- Reactive Business Approach – businesses realise there’s opportunity here to digitise what they already do today. Boards are generally slow and risk adverse (I’ve met exceptions believe you me) and want to wait till…. something else is in place before they act…usually in the 3-5 year plan.
- The response to the changing market, the API economy is all knee jerk and reactive from the top.
- The required security practices, systems and solutions agility never realised because they require forward thinking & upfront investment. Businesses that can do this, become real competitors.
- Business Inertia – “This is the way we’ve done it for the last 20 years…and this is the way we’re doing it for the next 20 years….”. I’m sure we all know a few of these.
Time is right for the era of the disruptor….
Just on that last point in particular, the climate is such that it’s ripe for the world of disruptive businesses/solutions to take hold and really cause a stir in the established industries. If we look at things like:
- Uber (of course)
- Many others (I’m working with several clients that will disrupt a couple of huge areas in business)
Are big businesses concerned? I see the ones that want to be ahead of the curve with a structure of something like:
- R&D unit – for things like ‘test & learn’ + innovation.
- Internal optimisations
- Seeding the Tech Start-up industry – sponsor, mentor & fund various startups. But leave them alone in their own ecosystem, rather than bring them into the larger well established business for fear of strangling their ‘mo-jo’
From an idea to commercialisation…. the time has never been better….