2020 was obviously a radically different year than its predecessors. This time last year, I was regrouping from my last airborne (mis)adventure of the year, via the Vegas enterprise event circuit.
No middle seats for me this fall, but one thing was a constant: more interesting stories than I've had time to write.
One that jumps out: a (virtual) panel of Constellation Research Supernova Award finalists, moderated by Constellation's Doug Henschen.The panel category was one of my faves: data-to-decisions (This panel was part of Constellation's first-ever virtual version of the Constellation Connected Enterprise event).
I guess you could refine it further, as in: data-to-good-decisions. Shiny tools may help. So will rolling up your sleeves, and getting that data plumbing sorted. AI and predictive capabilities? Probably. But: we still need to make good decisions, and act on them. So how have the panelists fared?
Shell.ai - deriving benefits from a multi-cloud approch
Henschen started with Dan Jeavons, General Manager of Data Science at Shell.ai (Shell.ai went on to win the Supernova award in this category). Jeavons has been at this for a while - he's a three-time Supernova Awards finalist. As Henschen explained, Jeavons and team aren't just dabbling in pilot projects. Their mission is to bring machine learning, IoT and predictive maintenance to large scale at Shell. To pull that off, Jeavons spent the last seven years building Shell's Advanced Analytics Center of Excellence. When Jeavons built Shell.ai's data platform, he made a big decision: a multi-cloud approach.
If anything, 2020 was the year the shine came off of the "multi-cloud" trend, as enterprises grappled with the complexities of cloud management underneath the multi-cloud hype. So Henschen asked Deavons: why multi-cloud? Jeavons responded:
It's been an amazing journey for us at Shell. We are cloud-first in terms of our strategy. So, we've been an early adopter of cloud technology. I was actually pushing clouds when we started this journey, way back in 2013.
As part of that, we recognize that Shell is a pretty big place. There's multiple different businesses. Actually, there are some real merits to different cloud technologies for different businesses, for different reasons. What we've tried to do is make sure we maintain that that edge, that multi-cloud capability, but also maintain the portability. We want to be able to deploy things across multiple cloud environments. So that's been really important to us.
But what about companies struggling to derive multi-cloud benefits? Jeavons advises cultivating expertise in specific clouds, placing those experts on the proper dedicated teams.
I think [multi-cloud] is helpful commercially. But it's also helpful from the pure technology side. What we've tried to do is make sure that we empower the people that are working in the environment. So trying to allow experts who are engineers in a particular platform to be able to make use of their skills, and these skills are scarce. So you want to help to enable those who have those skills to work in ways that they're familiar with.
The Global Emancipation Network - from data silos to predictive analytics
The next panelist hit on the challenges from a non-profit angle. Sherry Caltagirone is the founder and executive director of the Global Emancipation Network. Their mission? Fight human trafficking - from sex trafficking to forced adoptions and domestic servitude. It's a cause that shouldn't be part of modern life, yet the ugliness persists. Based on a conservative statistic of reported incidents through 2017, labor trafficking impacts 24.9 million victims, sex trafficking 4.8 million.
Caltagirone and her team are trying to stop these incidents with big data analysis. That's a daunting undertaking, with siloed databases everywhere. Pulling together data from law enforcement, government, and academic sources is not a minor project. As Caltagirone told CCE 2020 attendees:
We're talking about a problem that has between 20 million and 40 million victims. Now, I always like to point out, that's over 100% margin of error, just in that data point alone... So it shows that we're really in a nascent field when it comes to applying data and analytics to human trafficking. But for the Global Emancipation Network, we've always begun with the data and intelligence approach to it.
They've learned plenty about tackling data formats - and variety. Caltagirone:
The data comes in so many different formats. That's both a challenge, and it's really exciting - it keeps it interesting. Whether it's sex trafficking and brokering ads; it could be social media. It can be media articles - it really runs the gamut.
Getting that data plumbing right is essential. But it always comes back to: how can you act in meaningful ways?
Data is completely meaningless, unless it's machinable. Unless you can search for the correlations between it, unless you can translate that into actionable insights. For instance, we have a tool we developed with Accenture that's called Artemis that really helps us rapidly identify risk, whether it's risky businesses, individual content online and whatnot. [Then we can] triage the workflow, because of course, it's a resource-intensive effort to search through all of these data points, and try to find the needle in the haystack, right?
Another point many of us can relate to: looking after our own well-being, while dealing with tough problems. Caltagirone acknowledged the "mental burdens" of what they do:
It's something that can't be understated. Anyone who's working in this space, or any sort of child protection, it's really difficult to separate yourself from those works. So those are the things that we try to do a global emancipation network to convert all that data, correlate it, and share it out to the community.
Muscular Dystrophy Association - from legacy data sprawl to results
We don't have to go outside the organization to find data silo problems. That was the case with the next panelist, Michael Kennedy, CFO, the Muscular Dystrophy Association (Kennedy was actually a finalist in a different Supernova category, tech optimization, but the data theme fits in). Kennedy's goal? To bring financial visibility by centralizing and modernizing the MDA's data architecture. Another key: integrate the front and back office. So what has Kennedy's team learned from this 18-month project? Field lesson number one: eradicate legacy data sprawl. As he told us:
I'm very honored to represent MDA in this - it was just a tremendous effort by a lot of people. When I first came to MDA, I'm looking at a very antiquated back office and technology structure that becomes very costly. Like a lot of companies, when you don't invest in that back office and technology, you get a patchwork. We had over 65 disparate systems across 100 offices.
Field lesson number two: always keep the business outcome emblazoned in your mind. Kennedy:
The real drive was to try to get more of those dollars over to the mission. And so we just ripped everything apart. We just said, "No, everything we had, we don't want." And so it was a huge, huge change effort. And we got up and running within seven months with the core part of it. And then it took about another five months to get the revenue-producing pieces in there.
There's never truly an "after" anymore. But how do things stand today?
Now we're sitting with a Salesforce and FinancialForce platform that's cloud-based, that allows me to create centralized service centers, that allows me to reduce our footprint of those offices. And, you know, it's also saved at least at least $10 million on an annual basis. As a non-profit, I don't have an event that raises that amount of money.
And that may be the biggest takeaway from this panel. It's not just getting the data platform right. You want to change your role within the business. That changes your bottom line; it can also change your cultural buy-in. Kennedy:
You get to a point where you say "Okay, now the finance organization, the operation organization to start to contribute to the mission." And that starts building up a morale issue for you where people starting to feel really good about what they're doing.
My take - virtual events bring project know-how, but need a creative jolt
Tania Zieja, CFO of Halloran Consulting Group, spoke to similar challenges. Zieja led the push to modernize their back office, and integrate siloed data. I thought Kennedy captured these types of projects well: your transformation projects are "only as good as the innovation that comes on the back end."
For the MDA, that was put to the test during the pandemic. It's a story we've documented often on diginomica: transformations well underway prior to the pandemic proved themselves crucial. The MDA was able to convert to virtual fundraisers this year, aided in no small part by freeing up their talent from back office administrivia. As Kennedy says:
By allowing the revenue producers to stop worrying about administrative functions, they really got creative. We just pulled off a great event just recently.
Readers know I haven't had much love for virtual events in 2020. Instead of the creativity Kennedy referred to, we've seen a profound lack of it.
However, despite some qualms I have with the panel format, a live, unscripted customer panel like this one is always valuable - never more so than this year, where it's harder to learn from projects on the ground. The unscripted part of this matters; project use cases must pass the smell test. Unrehearsed/imperfect comments provide the most credible lessons.
Overall, CCE 2020 had an above-average virtual event experience. The pace of the discussions at CCE was speedy and entertaining, but to the point where sometimes, the surface was only scratched. Twenty-minute sessions work well for trendy topics like the future of virtual reality, but less well for pulling out project lessons. A customer panel like this can realistically go on for at least 45 minutes, perhaps longer, integrating questions from the audience and the social stream.
I give Constellation Research founder Ray Wang and team kudos for taking the risk of running a simultaneous social gathering platform alongside their streaming content. Few event organizers had the guts to try this in 2020; they are too insecure about drawing attention away from the pundits on the virtual main stage.
In Constellation's case, the platform they used for this at CCE, Wonder.me, was not ready for prime time. It wasn't easy enough to find other attendees for impromptu conversations. There was too much wander and not enough wonder. I've had better experiences with a similar platform, Gatherly, which I helped to moderate during a college alumni event. But Gatherly also had weaknesses, such as multiple floors which served to separate online attendees for no apparent reason.
In fairness to Constellation, these online mingling tools are in their relative infancy. I think most event producers and organizers are eager to fast-forward back to on-the-ground events and leave these online events behind. That would be a mistake. I was glad to see CCE take some creative chances on this event; I hope they continue to do so. Not everyone can make it to the secluded splendor of Half Moon Bay, even in healthier times.