Testing Industry Perceptions at Data Management Summit London

Every year at the A-Team Group Data Management Summit we take the pulse of the financial data and tech industry on a range of critical topics of the day. We do this through audience participation questions during the day-long event, urging delegates to interact with speakers and other participants via remote voting on salient questions.

It’s an effective method of divining what matters most to data and technology leaders in capital markets.

This year’s summit highlighted some fascinating insights into perceptions about subjects ranging from corporate culture, attitude to change and, of course, artificial intelligence.

Setting a Baseline

Interestingly, stakeholder engagement was a sticking point for a large number of respondents. Again, the day’s discussions would highlight this, along with corporate culture, as an impediment to innovation and change.

Another question set to establish a baseline for current thinking in the industry asked data leaders which functions take up most of their time. The response was unanimous – simply keeping operations running as usual was the biggest use of leaders’ days in the office.

Novelty of Data Products and AI in Data Management

Responses to a question during the first panel session reflected the relative novelty of data products and marketplaces. Asked where their organisations were on their data products journey, delegates were fairly evenly split between those that had scaled the products across domains through to those who had defined standards but hadn’t widely adopted products.

The results were positive, noted the panel, suggesting that the concept had come a long way in a short space of time.

In a new feature of DMS London, delegates were asked at the end of some sessions to list their future priorities with regards to the theme of each discussion. In the first they were asked to prioritise their aims for accelerating business value from data products.

Unsurprisingly, the leading ambitions were the most fundamental considerations; clarifying ownership and accountability came just above strengthening cross-functional collaboration. The former being a vital precursor to building out a data product strategy and the latter being important for ensuring the broadest effectiveness of products.

The first question to delve into attitudes towards AI asked to what extent is AI in data management a strategic priority critical to business success. The response was resounding, with most saying it was a the top priority with full backing and dedicated funding. The next largest group of respondents said they saw AI as a major priority but were still piloting programmes in a bid to understand business value.

Speakers returned to the essence of both responses again and again during the summit’s sessions as participants recounted their, or their clients’, initial caution over AI often quickly giving way to full-throttle deployment as the business opportunities of the technology were realised.

Taming Unstructured Data and Legacy Systems

Delegates were next asked if their organisations differentiated between structured and unstructured data. An overwhelming majority said yes, reflecting the growing importance of data locked in once difficult-to-reach sources.

At the end of a session dedicated to the subject, delegates were asked what they were likely to prioritise for improvement in terms of the discussed theme. Again the result was resounding with the largest proportion saying they would focus on building a unified governance framework across all data types, an essential part of deriving value out of both structured and unstructured data.

That touched on a theme discussed by the panel; can a governance framework be created for both data types? Participants agreed it would made sense but also accepted that it was yet to be seen broadly across the industry.

The challenge of overcoming the ages-old problem to fragmented legacy systems was given its own panel discussion. Unsurprisingly, when the next question was asked – “What is the biggest barrier to building a unified data ecosystem in your organisation?” – the leading response was legacy systems and technical debt. However, close on its heels was the response, “lack of clear data ownership and governance”. This fundamental ingredient in any data management shift is a recurrent theme in industry discourse as AI encourages organisations to reassess their business models.

Data lineage is an essential part of any data management strategy and the panel discussion dedicated to the topic asked delegates how mature their organisations’ data lineage capabilities were. The response to that and to the session’s question on post-session priorities, yielded complementary results. A large majority of respondents said their lineage capabilities were partially mature and focused on specific systems or domains. Reflecting this, the audience said it would prioritise integrating lineage with governance and controls.

Data Quality and Lineage

The quest for good quality data has taken on added importance with the widespread adoption of AI models, the quality of whose outputs are directly related to the integrity of the data fed into them. In the final panel discussion of the summit, a plurality of responses emerged when delegates were asked where their organisations are in their data integrity journeys.

The largest group said they were primarily reactive, utilising dashboards and issue logs, and a large minority said they were similarly responsive although they had automated rules and controls. Reflecting the early stages on the path to being able to prevent integrity-related issues, respondents to the final question – “Which will create the most competitive advantage over the next three years” – a majority voted for “faster remediation”.

Closing out the discussion, the panel commented that fast remediation was in itself a tricky task as it would require all parts of the enterprise to be behind any changes and also because it would depend on having full trust in their data.