We're excited to announce the release of our 30-Day Free Trial — check it out!

The CDO Magazine Summit was held in Cincinnati on September 12, 2023 and it was a huge success.  In its fifth year, the Summit and its publisher Steve Wanamaker gathered over 500 Chief Data Officers, Chief Analytics Officers, as well as VPs and Directors of AI, Analytics, Business Intelligence, and Enterprise Data Management to talk about the latest trends and technologies.  Gen AI was mentioned in every keynote and breakout session.  Other recurring topics included the CDO’s role in the enterprise, modern data stack ROI, data literacy, and change management.  

I attended the summit as an Advisor to Spectio, a business intelligence workflow and collaboration startup, but also as a public company board director (Ooma, a leading VoIP telecom company) and as a B2B SaaS CEO.  While at the event, I felt a sense of deja vu going back 15 years to some of the first big digital marketing technology conferences, when I was CEO of a marketing technology and analytics company.  Back then, “CDO” was the newest title in the C-suite, but referred more to Chief Digital Officer than Chief Data Officer.  Though theoretically part of the C-suite, Chief Digital Officers generally reported to the CMO and were of course wanting to report directly to the CEO.  Google was the big disruptive technology, so CEOs asked every CIO or CMO, “What’s our Google strategy?” rather than “What’s our digital marketing strategy” or “What’s our search marketing strategy.”  The CDOs of the time were refuting the other famous Wanamaker (John Wanamaker), who infamously said, “I know half my advertising is wasted, I just don’t know which half.”  Chief Digital Officers suddenly had very detailed metrics around marketing ROI and the marketing technology stack exploded with platforms, tools and services for every type of media: search, display, social, video, mobile.  Forrester and Gartner started entirely new practices covering cutting edge startups in the marketing tech stack and technology enabled service providers.  VCs and investment bankers also developed entire practices around these exciting companies.  All of these things are happening again, now, for the new CDO — the Chief Digital Officer — but with GenAI, the modern data stack, and the entire IT industry at an important inflection point.

The 2023 CDO Magazine Summit covered a wide range of topics, with over 30 incredible breakout sessions.   Here is the summary of what I heard from the keynotes and sessions I attended:

  1. GenAI was the hot topic:  From the very first keynote featuring Tonjia Coverdale at Associated Bank and Jasmine de Gaia at Wells Fargo, generative AI was top of mind.  Almost all CDOs — across industries, from banking, healthcare, retail, manufacturing and others — are being asked by their CEOs and boards about GenAI.  There is a lot of excitement around the opportunity, but also just as much concern about privacy and security risks.  FOMO (fear of missing out) has a high degree of tension against the FOLS (fear of looking stupid) or FOGS (fear of getting sued).  Many CDOs took pains to mention that GenAI is just the shiny new object, and that many enterprises have been deploying other types of AI and machine learning capabilities for years.  Dan Merzlyak from BlackRock made an amusing observation that, “Everyone wants to work on and use AI, but no one wants to take the responsibility for it.”  — and CDOs need to be the people who do.   
  2. GenAI opens the door to budget around the entire modern data stack:  CDOs are using the excitement around GenAI to educate and expand budgets to more broadly fund the entire modern data stack that encompasses data collection, data storage, data processing, and data analysis.
  3. Quantifiable Data ROI is a requirement, and CDOs are ready:  CDOs and CDAOs are all highly attuned to the holy trinity of ROI: revenue uplift, cost takeout, and risk mitigation.  Every speaker mentioned that each data initiative they deploy is geared to improve at least one if not all of these drivers of success.  They took pains to differentiate from the science projects of the past, which required lots of money, lots of waiting, and lots of belief with little evidence of payback.  Rob Golden from the Great American Insurance Group described how finely tuned every data and analytics project was to profitability KPIs, and how he could prioritize projects and budgets based on the ROI impact.  Most speakers told stories of how their machine learning models and analytics models were able to increase customer acquisition in the double percentage digits and reduce service call times / labor costs / sales incentive costs.
  4. CDOs are still reporting to CIOs and not really part of the C-Suite:  CDOs and CDAOs are for the most part still reporting to CIOs, who may still be reporting to CFOs.  Just like in the past, when Chief Digital Officers wanted to be equal to CMOs and report into the CEO, today’s Chief Data Officers are also wanting a “seat in the suite” and report directly to the CEOs.  Somewhat interestingly, like the CMOs and the Chief Digital Officers of the past, the average tenure of the new Chief Data Officer is only 18 months.  Sue Pittacora of Wavicle Solutions quoted a recent HBR article noting that “only 35% of CDOs are successful.”  Apparently, with greater visibility comes greater accountability and expectations.
  5. “Who gets the credit” and cost center vs profit center models are a concern:  Tying together point #3 about ROI and point #4 about reporting and role of the CDO/CDAO, there were some recurring themes around “who gets the credit” for success and if the data organization is a profit center or a cost center.  If the data organization creates a targeting model that achieves a 7% uplift on field sales efforts, does the sales team get the credit or does the data team get it?   Who gets more budget?  If the data team is a service organization (read: cost center) for their business partners and internal stakeholders, then the sales team takes the credit.  If the data team attempts to set up like a profit center, they may act more like an internal vendor and charge each business unit or function like an internal customer, with internal transfer pricing or cost shifting.  Dipti Desai from Global Payments noted that being a trusted advisor to internal business partners and acting as a service organization appears to be the best approach, while agreeing with those partners to share the success metrics and “double dip” on the credit.
  6. Operating models for data governance are still in flux:  Another question that many data executives kept asking each other was around the best model for data governance: centralized vs federated.  One speaker said “anything less than centralized is chaos,” noting that “federated models create skunkworks, which end up defeating the purpose.”  On the flip side, some leaders noted that in order to get buy-in and true utilization from every global business unit and function, that governance needed to be more lightweight, collaborative, and flexible.   There seemed to be no best answer for this question, and best practices are evolving.
  7. It’s not just about the tech because data literacy and upskilling are required across every enterprise: One of the more interesting recurring topics was the need for data literacy, data training, and upskilling across the entire enterprise, from the front lines to the executive levels.  Iwao Fusillo of Pepsico gave a great talk about “democratizing analytics.”  Unlike other functions, like marketing which is primarily left to the marketers, data usage and activation occurs at every level and at every function in a company.  But even with all that data and tooling, if business stakeholders don’t know how to ask the right questions of the data, are not sufficiently equipped to use business intelligence tools, or aren’t sure how to analyze and interpret information — all the investment isn’t worth anything.  One industry analyst suggested that enterprises only get about 25% adoption of their business intelligence solutions.
  8. Deduplicating and decommissioning data, dashboards, and data warehouses are harder than it seems:  One of the more surprising items that was mentioned multiple times from the stage is the high cost of deduplicating and decommissioning data, dashboards, and data warehouses.  With data spread across on-prem, hybrid, and cloud based systems, along with decades of reports and plumbing without any documentation — the storage and compute costs are immense, but just “shutting things down” is not an option.  Jasmine de Gaia also mentioned taking 18 months and a whole change management team to consolidate 5 data warehouses into 1, having to move thousands of end users over in waves, and being nervous up until the very end of the decommissioning on whether something would break, but doing it all very successfully.
  9. GenAI hallucinations, lies, and output quality monitoring are a serious concern:  Returning to the GenAI themes, one the concerns voiced multiple times was that ChatGPT and other LLMs generate hallucinations, and in fact will cite real-world sources with fake statistics, including realistic looking URLs that only point to 404 pages.  The authoritative way that LLMs present information creates the illusion of quality that is not there.  While people more involved with AI initiatives are aware of these issues, many more executives and front line workers are not aware of these limitations.  Without training, tools, and context to closely monitor GenAI inputs, pipelines, and outputs with discernment, the potential for damage in all kinds of AI based deployments is high.  Shradha Puntambekar of American Modern Insurance Group stated plainly, “People on the business side need to be able to give AI the sniff test.”
  10. Ethical AI, unbiased AI, and DEI in AI need more attention: One of the less obvious but most important comments came in the opening panel with Tonjia Coverdale, and was reiterated in various ways in breakout sessions  In particular, Neelima Sriramula from Southwest Airlines noted, “All CDOs and CDAOs need to be constantly aware of the ethical challenges that genAI and all AI might create, particularly around DEI and model bias.”  From making offers for consumer credit or loans, to evaluating candidates and employees for hiring or promotions, to legal document reviews and academic intervention models, almost every use case for any kind of AI may have significant repercussions if diversity, equity and inclusion are not monitored closely.  Many studies and stories have already been published about various LLMs and AIs displaying biases and discrimination.  Scott Peachey from Invesco noted that “As executives, we’re all mandated to get trained and pass tests annually on business ethics, anti-discrimination and anti-harassment.  Should all AIs have to pass an annual ethics test as well?”   Yet, the tension remains – when a CEO, CFO or CRO is driving for results and is afraid of being left behind by competitors — it’s easy to say “damn the torpedoes” and press forward.  Circling back to point #1, everyone wants to leverage GenAI, but no one wants to take the responsibility for the repercussions.   It was great to hear that the CDOs at the CDO Magazine Summit were acutely aware of this tension and actively discussing how to balance the opportunity and the risk.

Despite the concerns raised in the last few points, the overall vibe at the conference was one of excitement and opportunity.  The CDO/CDAO role is relatively new, and coupled with GenAI having its “moment,” (just as with the advent of Google, iPhone, Cloud, or 5G), data is taking an exponential step forward in a way that will transform businesses and our lives.  Attending on behalf of Spectio as a sponsor, the conference had high ROI from a networking and learning standpoint.  As a board director and seasoned CEO, the conference was highly enlightening and fascinating to me on multiple fronts – the opportunities and risks, the innovations and the requirements around AI and data, but also the people and processes required to be successful.  More than anything, I came away hugely impressed with the professionalism, competency and thoughtfulness of this generation of CDOs and CDAOs.  They recognize that this is their time to shine, innovate, lead and succeed, but do it in an ethical and very intentional way.  Boards and CEOs should definitely pay attention to their CDOs and CDAOs to help their organizations succeed in the next wave of transformation.

Spectio is an AI-Powered PRD Solution for Data, Analytics, and BI. It automates requirements gathering, facilitates dashboard prototyping, and maintains documentation for datasets, dashboards, and reports. Spectio is designed to address the business-data translation gap to drive better outcomes.

Sign Up for Spectio
Others You Might Like
No items found.