Enterprise hits and misses - data science meets project reality, extreme heterogeneity sparks debate, and TikTok awaits resolution

Profile picture for user jreed By Jon Reed September 28, 2020
Summary:
This week - applied data science brings analytics project lessons, while enterprises struggle with "extreme heterogeneity." AI and automation brings a white collar wake-up call, and diginomica finally weighs in on Oracle-TikTok. We also wind up in the whiffs section.

King Checkmate

Lead story - Applying data science to real world data projects - a use case roundup

MyPOV: It was an analytics use case type of week at diginomica. Derek got the action started in There's no data science unicorn - building a data team at HSBC s The big lesson on scaling data science? Quit scrounging the applicant pool for unicorns - build diverse teams instead. Derek quotes HSBC's Global Head of Analytics:

Boteju said that he's realised during his time working in the field of data science that there is "no such thing as a data unicorn". He explained that four or five years ago there was a view that a single person could do everything - from data engineering, to building algorithms, to communicating really well and pulling it all together. However, this is not true and for real scale, you need to build a diverse team of talents with individual skills and knowledge.

Drilling into industry is where this needs go to next. As in Derek's Using data to reimagine the real estate industry at Zoopla. Via the Big Data LDN event this week, Derek relays lessons from Zoopla on their goal of company-wide data literacy:

The era of the algorithm economy - augmented analytics is one of the things that will come. A lot of jobs, a lot of the things people are doing right now, will be automated, whether we want it or not. You need to understand how fast this is evolving.

Yes - the pace of "general AI" may be slow, but automation is not - a wakeup call we'll return to momentarily. For a graph database angle, check Gary's Allianz - we’ve made €2m profit using graph technology.

Diginomica picks - my top stories on diginomica this week

Vendor analysis, diginomica style.

The Oracle TikTok affair - Yeah, we finally added our take to the TikTokkery. Stuart did the needful last Monday in TikTok, Oracle, Walmart and Trump - the shape of things to come?(Updated)?.This story continues to stall out and create more analyst hair loss evolve. We're not planning to go skinny dipping for page views add to the noise, but when the story advances, we'll be there.

My top vendor picks this week:

Inforum 2020 is in the books, but the customer use cases rolled on:

SAPfans and ERP watchers got another diginomica dollop of SAP developer drama, dialogue and use cases:

A few more vendor picks, without the quotables:

Jon's grab bag - Derek parses the annual Harvey Nash-KPMG CIO survey in How COVID-19 has changed - or not changed - CIO priorities. Gist: security and automation investments are in - financial applications or systems of record? Not so much.

Meanwhile, Neil bears down on the AI explainability apologists in The explainability problem - can new approaches pry open the AI black box? - and raises some interesting ways forward.

Best of the rest

Waiter suggesting a bottle of wine to a customer

Josh Greenbaum doesn't blog often these days, but he can be counted on for a creative/provocative take, as in: Customer Success, Vendor Empathy, and the Problem of Extreme Heterogeneity. I don't agree with all the points in here - for example I don't think suite versus best-of-breed is the best juxtaposition right now (I think it's more about aligning with an applications platform and integrating from there). But agreement isn't the point. Greenbaum puts enterprise software incumbents on notice:

What's important for the vendors that sell enterprise software, particularly the suite vendors – as their best-of-breed counterparts have understood the heterogeneity issue for a while – is that if they don't figure out how to do a better job talking to their customers about supporting other vendors' products, including competitors, they'll be co-opted by the hyperscalers.

Such a development would, of course, be a mockery to the "customer success" marketing programs practices these vendors have adopted. Greenbaum didn't miss the irony. I don't think the "extreme heterogeneity" Greenbaum describes here is a viable way forward, but he's right about the reality of it, in most $1 billion+ companies.

Honorable mention

Overworked businessman

Whiffs

Brexit - your go-to source for tragicomic IT snafus: Brexit travel permits designed to avoid 7,000-lorry jams come January depends on software that won't be finished till April. Including the dire-but-priceless line from The Register:

A Cabinet Office official told Bloomberg that beta is a standard labelling practice for a digital service that is fully operational. Experienced IT professionals may contest this definition.

Oh yeah, about those indoor Amazon drones:

And that rough week in IoT security I referred to earlier:

Finally, let's keep those hot mics off - especially when diginomica is around.

See you next time...

If you find an #ensw piece that qualifies for hits and misses - in a good or bad way - let me know in the comments as Clive (almost) always does. Most Enterprise hits and misses articles are selected from my curated @jonerpnewsfeed. 'myPOV' is borrowed with reluctant permission from the ubiquitous Ray Wang.