Digital Analytics Hub 2019 Recap
Incredibly, it is already four weeks since the DA Hub 2019. Running a conference means work continues well after the last delegate has left the venue. But now that we have had time to reflect, we wanted to share some of our highlights from the past DA Hub.
First, a quick, but profound, thank you to the folks who made this event such a success - Rafa, Grace, and the incredible Lansdowne Resort staff. While we’re still analyzing the feedback we received over the course of the conference, we already know that the 2019 conference was, hands down, our best event to date. We look forward to deep diving on the feedback over the next few weeks and planning to make 2020 even better.
Without further ado, here are some of the highlights...
UnSummit, run by industry stalwart David McBride, has been a staple of the pre-Adobe Summit events for the past ten years. Being a big fan of the DA Hub, David agreed to bring the format to the conference for the first time. Based on the great feedback we’ve received, we plan to make UnSummit at DA Hub a continuing thought leadership and experience-sharing event for the most senior management in analytics.
This year’s UnSummit included presentations by senior leaders from Capital One, NY Times, NP Digital, Softcrylic, Conductrics, Neebo & SynthetIQ, each followed by a short Q&A. Presenters shared a variety of use cases, learnings and fresh ideas in analytics and experimentation. Some highlights were…
- Matt Gershoff’s (Conductrics) brilliantly detailed, yet layman-friendly, review of multi-arm bandit methodologies in experimentation (#mattgershoffed)
- Karen Bellin’s (NP Digital) segmentation framework
- Bob Page’s world premier introduction of the Neebo platform, a cloud solution that unifies analytics across enterprise applications, reporting tools, and data silos
While totally different from the huddle format DA Hub is famous for, the UnSummit was well received by our audience as a great way to get exposure to new practices in the industry. We are already working with David to increase the value derived by attendees.
We were lucky to have three of the industry’s biggest names lead workshops this year. John Lovett (Search Discovery) ran a thought-provoking Data Strategy workshop. Amy Wilder (EY) shared invaluable insights on how to become a valuable analytics consultant, internally in your organization or with clients. And Anil Batra (Optizent) ran two great SQL for marketing analytics workshops. We’ll be sharing more on the workshops next week, so stay tuned!
Huddles – the DA Hub Main Show!
The heart of the DA Hub are the amazing huddle conversations, and we kicked off with the heavy weights round. Every year we choose the most experienced leaders to run the first-round huddles to get us off to a great start and to allow our new huddle leaders to experience the best leaders prior to their own huddles. With leaders such as June Dershewitz (Amazon), Rusty Rahmer (GSK), Paula Sappington (Hilton) and James Robinson (NY Times), you know you’re in for a great discussion and knowledge exchange. Here are a few highlights but hop over to our LinkedIn group for comprehensive huddle coverage.
- June Dershewitz (Amazon) Data Culture & Self-Service huddle – building on her incredibly popular self-service huddle last year, June returned with new insights on the topic. The group discussed the meaning of data democracy vs data literacy in the organization. It defined the funnel for self-service success. But most importantly, came up with a list of practical suggestions to try out (find them here)
- Rastko Kovacevic (Delta Air Lines) Personalization huddle – Customer focus should be at center of all personalization efforts. We should always ask ourselves whether we’re doing something good for the customer. The discussion also highlighted the differences between content personalization and ecommerce personalization. The former is more acceptable and viewed as customer–focused. The latter should be carefully executed to ensure value to customer is notable
- James Robinson (NY Times) Analytics on Analytics huddle – attending James’ huddles is always a delight. James is a passionate leader and always well-prepared with a clear huddle plan. In this huddle we discussed the importance for analytics teams of not only understanding stakeholders’ needs but also to be able to monitor and analyze the usage of the team’s output for those stakeholders. Most participants expressed degrees of unease/uncertainty tracking emails or dashboards to see if those were being used internally. But we concluded that, done openly and responsibly, tracking internal use would be incredibly useful for utilizing analytics resources more effectively and in defining the value of the analytics team
- Paula Sappington (Hilton) Optimizing Your Optimization Program huddle – this is another huddle that got excellent feedback in 2018 so we asked Paula to run an updated version this year. The discussion initially focused on qualitative research in experimentation. The group agreed that it is important to do both types of testing – everyone had stories of qual testing that then resulted in a different outcome when A/B tested. The discussion went onto testing prioritization. While many challenges still persist with prioritization, the group was able to offer some solutions to help meet those challenges
- Sid Shan (Condé Nast International) Applying ML in Digital Analytics huddle – the group discussed use cases of ML in analytics process and how data modeling can support business process. Data audit was identified as one of the simplest use cases of ML in the analytics process. There was consensus that before implementing ML, data leaders must be clear on real business problems data models will solve. Unfortunately, often that clarity isn’t there. Explainability (defined as the extent to which the internal mechanics of a machine or deep learning system can be explained in human terms) or articulating data models in simple business language, is a critical success factor. ML project lifecycle may be highly iterative – determine project feasibility; built/refine data models; revisit feasibility and get model to production
- Rusty Rahmer (GSK) Analytics Team Org Model Best Practices huddle – the group agreed that analytics teams will continue to expand (new roles), deepen (in expertise and complexity) and increasingly include partners to support the consumption of data and analytics. Consequently, how the business organizes is less important than embracing it intentionally and purposefully. Communication was identified as the biggest factor for future success engulfing many organizational aspects from structure through hiring to collaboration tools
- Tom Betts (FT) Privacy huddle – GDPR is driving global interest in data protection/privacy. Companies have mixed experiences of how to implement with different levels of maturity. New laws and guidance from regulators makes this a dynamic domain with little best practice. Establishing processes and right culture deemed more important than the technology
Tête à Tête– Business Speed Dating
The Tête à Tête is a pretty unique feature to the DA Hub. It is an opportunity for client-side practitioners to have a quick chat with industry consultants and vendors but without being sales-focused. Instead delegates get an opportunity to discuss their challenges and hear about potential solutions. The rapid nature of the track (10-minute meetings) means you cannot get drawn into a long boring conversation. We arranged 95 meetings over the course of one and a half hours. Delegates constantly tell us how valuable these meetings are for them.
After hours networking
More networking time with industry peers was a common delegate request over the years. We were attentive but the original two-day conference format simply didn’t allow the time for more networking events. Not wanting to reduce the number of huddle rounds, we decided to expand the conference to three days.
We must admit we were concerned delegates wouldn’t be able to afford three days out of the office. But as it turned out 90% of this year’s delegates attended all three days (we also offered two-day passes). That offered all of us more time to continue those great conversations from the huddles.
We concluded the first conference day with a fantastic sunset dinner at the outdoors Pavilion. Overlooking one of the Lansdowne’s three pristine golf courses and adjacent to the teeing green, the pavilion offered the perfect settings for a sociable meal and drinks for the entire conference audience.
Following the meal, attention shifted to a friendly (yet passionate) cornhole competition. Loudoun County in Virginia is famous for its alcohol production. So together with
The second day was the ‘short’ day with the afternoon dedicated to networking activities or free time. Kelly Wortham and Search Discovery hosted a TLC Happy Hour following a live panel Q&A on Analytics for Experimentation. Situated on the Lansdowne’s main terrace overseeing the golf range with the Potomac river in the background, it was the perfect event to bridge between a busy conference day and the gala dinner that evening. The peach sangrias were the afternoon stars of the show. We finished the day at the spacious golf club with a luxurious South American dinner buffet followed by a networking reception hosted by Conductrics. Then a bunch of delegates continued to the hotel bar for the traditional mid-conference night cap gathering.
For our final social event we returned to the terrace following the conclusion of the final day. It was a time to take in the last rays of summer, enjoy some reflective thoughts on the past three days and bid farewell to our brilliant delegates that make the DA Hub the incredible gathering it is.
We have created a new DA Hub LinkedIn Conference page to accompany the existing Group page. The Conference page will be used for updates about the conference. The Group page will continue to act as the huddle summary repository.
We are looking at dates and venue for 2020. We expect to have those details available by early November 2019.
Matthias and Michael