The false economies of data team cost-cuttingby
Paul Laughlin unpacks some of the common mistakes made by businesses when trying to cut costs from the data team, that end up doing more harm than good.
In my two previous articles - How data leaders can cut costs without reducing their contribution to the company and Reduce, reuse and recycle: How efficient is your data usage? - I focused on what leaders can do to cut costs and drive efficiencies. But now it feels like time to call out some potential missteps.
These are the three mistakes that I have either seen other leaders made, or that I have made myself. The circumstances in which these occur normally comes from a business that is feeling under pressure and failing to see the value in what they suggest is curtailed.
Below I will outline why I believe each of these ‘opportunities‘ for cost saving often prove to be false economies: promising savings but actually either incurring greater cost in future or reducing benefit delivered even more than any cost saving achieved. I hope you will agree each is worth avoiding or challenging if you are under pressure from stakeholders to take that action.
Let’s outsource to avoid all these staff costs
One of the popular targets of cost cutters when they see larger data, insight or analytics teams is the number of employed staff. For many businesses their largest cost line is staffing. So, it is natural for them to review the potential to achieve their goals with less staff. Plus, finance leaders will be attracted to the potential to “variablise costs” (covert fixed costs into variable) by use of outsourced on-demand services instead.
Despite the high costs of data scientists and data engineers within data teams, they are sometimes protected from this by their mystique. Even business leaders have heard such roles are hard to fill and talent is in demand. The most frequent target I have seen is market research teams, closely followed by those working on data quality, data management or others who have not been given sexy new job titles.
To help stakeholders understand why this is a false economy, experienced data leaders need to proactively bring to life the value that such teams add. I have shared before on the need to both raise awareness of the value that internal researchers can add and protect technical backroom roles.
It can also be worth data leaders sharing how extra costs can be incurred and lower quality delivery when just buying in these services. Any leader who has tried this should have war stories of rework caused by lack of domain knowledge and reinventing the wheel each time.
Stop buying data, make do with what we’ve got
Another large cost line often spotted by the beancounters when reviewing analytics teams is the cost of external data purchase. So, conversely to the last challenge, data leaders can find themselves under pressure to end these costly external contracts. Once again, at face value this can make sense, particularly to organisations with large internal databases on clients etc.
Decisioning or database marketing teams will often be the chief casualties of this cost cutting idea. The nature of their work can require all types of external data (prospect pools, suppression files, data enhancement/imputation sources). It is also true that in the past these contracts have often been poorly managed/challenged, with suppliers appearing to be profiteering.
To understand how this can prove to be a false economy, data leaders need to point others to the often hidden costs. These can include reduced volume for marketing activity, risk of regulatory fines, level of returns or ‘goneaways‘. In fact those are the tip of a less visible iceberg, as lack of such data reduces data quality and can lead to misleading models and decisions.
The best tactic I have found here is tougher negotiation with data suppliers. Insist on matching exercises to ascertain data needed. Require full transparency to original data sources and timeliness. Refuse or reduce minimum usage targets, all to visibly shrink cost.
We can’t afford training or recruitment right now
A recruitment freeze is a very common reaction in businesses that need to better control costs. In some ways it is a very sensible way to quickly pull up the drawbridge to avoid any more raiding of limited funds.
However, a blanket non-negotiable edict rarely makes for wise management of the situation. Data leaders should not be put off from making the case as to why there may need to be exceptions.
The need to backfill staff who leave is felt particularly acutely in data teams, especially data science or data engineering teams. This is because of the high turnover of staff in such specialist functions. People who have grown or led such teams will know that they are always looking at opportunities elsewhere and how to keep developing.
With the war on talent still underway, businesses who leave specialist staff to cope with covering the extra work of others who have left, will soon lose them too. Demand is too high to get away with expecting such goodwill to last.
In a specialism that requires continual learning and attracts staff to stay with their current employer because of development opportunities, this approach is nuts.
Cutting training and other skills development resources really is an example of cutting off your nose to spite your face. I recommend that data leaders are proactive about educating their senior leadership on this issue. Ideally, prior to any recruitment freeze they need to be sharing both examples of value-add from team output and the rarity of such skills in today’s market.
If they do lose staff, they need to insist on radical prioritisation to reduce workload. Do less but well, rather than overburdening data staff and so losing more.
Which of those false economies have you challenged?
How did I do? Did you recognise at least one of those challenges in your business at the moment? If you have faced one of the above challenges, how did you effectively counter it (if you did)? Other, better, ideas of how to expose such false economies would be most welcome. I will gladly share to help our community.
Beyond these three posts, which other challenges have you faced during this global economic downturn? How have you (if you have) been asked to reduce costs and what has worked as a solution? Please let me know and I will gladly share the best ideas to help our community through these tight times.
I work with exceptional leaders & their teams, so they can master the people side of data & analytics.
That means helping them maximise the value they can drive from using data, analysis & research to intelligently interact with customers. It also means developing teams & enabling them to sustain their improvements through...