What a Chief Data Officer does for a Startup
It seems like there’s always a new kind of Chief Officer. A CDO is a “Chief Data Officer”; but why are startups finding they need them, and how did they manage before?
A startup’s data responsibilities
In every interaction your business has with suppliers, customers, regulatory bodies, the natural environment, and even itself, your organisation will handle data.
The data you collect will fall into various broad categories:
Transactional data are the ever changing bread-and-butter of your operations; current orders being processed, stock levels, employment records, bank balances, work in progress, and so on. They are usually fast-changing; not that any particular datum might change very often, but there’s lots of change afoot and any changes must usually be handled near-instantly.
Analytical data are historical records, generally kept as a log of transactional data as it was, or of changes to the transactional data. They are used for analysis - to study past behaviour, in order to try and predict future behaviour, by understanding the things and people inside and outside the business. Often, the analytical data is allowed to lag behind reality a bit, to enable cheaper “batch processing” mechanisms to be used - often by up to a day with modern systems, but once upon a time, quarterly often sufficed.
Reference data are “useful facts” kept for reference; usually either imported from some external source (eg, a supplier’s catalogue) or agreed internally (eg, your product categories). These are usually changed more rarely, often on a schedule such as when a new catalogue is published, but they need to be distributed to all parts of the organisation that need them in a timely manner.
You don’t need a CDO for all of that to happen, though. Teams will find the data they need, and set up infrastructure to store and process it. So, what value does a CDO bring?
The benefits of having a Chief Data Officer
As discussed in our post on data challenges for growing startups, the organic development of data systems and practices doesn’t always go well.
Startups need to ensure that legal and ethical data-management standards are upheld so there aren’t any legal timebombs ticking to come up in due diligence - and to avoid losing the confidence of users; and a big part of that work requires cross-team collaboration. When a GDPR request for a user’s data comes in, somebody needs to be sure that all the requested data about that individual has been found - so they need to know all the places where data about people might be. Insurers, investors, and business partners may require formal statements that certain data management standards are met, as part of due diligence checks and somebody needs to take board-level responsibility for signing that off.
Organic development of infrastructure is rapid and lets every team choose the best tool for the job, but it can also be inefficient. Each team buying their own licences and maintaining their own platforms, and having to set up adapters between platforms when teams need to share data, quickly adds up in cost. While a startup CDO shouldn’t be forcing teams to wait for central approval to build things, they should definitely be involved in tool-choice discussions so they can identify opportunities for sharing and standardisation, to avoid technical debt building up in the startup’s data infrastructure.
Without overall guidance, an organisation won’t develop a coherent culture around data. A CDO can set standards and policies - even informally - and educate the teams so that everyone, without needing to become a data specialist, understands enough about data to do their job properly. Everyone can then be confident their part in the organic development of the data infrastructure fits into a synergistic overall policy without needing the CDO to hand-hold them through everything. It’s better for everyone if the CDO can educate people and then delegate data decisions to them while being available for consultation on anything thorny and providing light-touch regulation, rather than being single-handedly responsible for everything data-related.
Having a big-picture view of where data is within an organisation and how it flows doesn’t just let you spot inefficiencies, it also lets you spot whole new opportunities. For instance, Usama Fayyad, CDO for Yahoo!, helped them grow their revenues from user targeting by 20 times in 4 years; and for a startup it’s even more important to find novel revenue streams, understand your users’ behaviour, and turn hunches into quantifiable key performance indicators.
So, does a startup need a CDO?
To begin with, certainly not. On the other hand, if you wait until the time your startup has a growing user base and is proving itself in the market, then your new CDO is going to have a big pile of research and technical debt clearance work to deal with before they can be effective!
In practice, a startup’s data management workload starts small, and grows; sometimes smoothly as data infrastructure organically grows within the organisation, sometimes in big chunks when regulations like GDPR need to be addressed before systems are opened to customer data. Somebody needs to address those needs, even if you don’t have a dedicated CDO on the payroll.
If you are lucky enough to have the required skill set within one or more of your existing team members, they can take it on - as long as they are given the time and resources away from the other demands of their job to do so! However, if nobody has those skills, or they’re too busy with everything else to keep up, we can help and support you via CDO as a service. This will enable you to plug skills gaps and build data teams easily, work with a senior data expert (with CDO experience) to implement data projects, and implement training workshops and programmes as needed, to build data literacy and up-skill staff quickly.
To learn more, see our article on CDO as a Service: The next data evolution or contact us for a free no obligation call to discuss your specific challenges. Talk to one of our Startup Specialists or Expert CDOs now.
Author
