Developing that vital Domain Knowledge in your analysts
How do you develop domain knowledge in your analysts, so their data usage, interpretation & recommendations make sense?
I’ve mentioned in a previous post, about the difficulties of offshoring analytics, how vital domain knowledge can be. Yet, I find most articles or speakers focus on your need to develop new technology skills (like mastering the latest ‘en vogue‘ coding language).
Akin to the greater importance of softer skills in analysts, as outlined previously, my own experience is that domain knowledge makes a greater difference to analyst effectiveness.
So, as my contribution to redressing that imbalance, here are some thoughts on domain knowledge. Why it matters & how best to help your analysts learn about their domain.
Domain Knowledge, what it is and why it matters
Let’s start with a definition. The Dictionary of Computing, usefully defines domain knowledge as:
“That knowledge which is specific to an application, as distinguished from general strategic or control knowledge that is independent of the details of any particular application. For example, data about the flight routes covered by a particular airline is domain knowledge, unlike search algorithms that might be used to locate the cheapest entry.”
Relevant to Customer Insight, Data Science or Analytics teams, I would emphasise this is knowledge about the real world context or business problem. Not just general awareness of principles (from business studies), nor just the technical knowledge of data/methods/technologies, but an understanding of what this means.
For example, an analyst might be fully familiar with data warehouse & coding skills required, but when asked to analyse customer behaviour when using a new channel (like mobile) may not have the knowledge required. Although they may be able to complete thorough descriptive analytics or other statistical work, without an understanding of customer context and what business is trying to achieve, they are unlikely to produce useful recommendations.
All too often I have seen apparently robust analysis be totally ignored in wider business, because it appears “naive” or “impractical“. Lacking an understanding of business strategy, goals for products/channels, current commercial performance/issues, customer needs & a “feeling” for the customer experience – can all lead to inaccurate interpretation of data items or irrelevant recommendations.
I’m sure other Customer Insight or Analytics leaders will concur, that such knowledge is key to both understanding the real business problem/need and making sensible recommendations as a result of analysis.
So, if such context is critical, how can analysts gain the knowledge needed. I will suggest 3 ways that I have seen prove effective.
Knowing your numbers, the key role of Commercial Understanding
Many an analyst or analytics manager has been ‘encouraged’, normally in their annual performance review, to become more ‘commercially aware‘.
One on the problems with such feedback is the vagueness of that phrase. What does an analyst need to know to be ‘commercially aware‘? From my own experience, I’d suggest there are 3 key dimensions of commercial knowledge:
- Strategy: Understanding the business strategy & why current participation (in products/channels/brands/segments) makes sense
- Competitive Market: Understanding who are the key competitors, the threat they pose & keeping up-to-date with relative performance & innovations
- Financial Performance: Understanding the key ‘profit drivers’ amidst commercial numbers, how to read financial reports & up-to-date on KPIs
Of those three, I find most businesses make an effort to share 1 & 3, but only at a high-level (suitable for staff at all levels). Analysts & analytical teams need a deeper understanding of strategy, assumptions and rationale for key decisions. Those insight or analytics teams working within Marketing teams will also probably see regular information on competitors & performance against market. This is especially true if they have followed my advice and brought research team into the same department as analytics teams (and ideally the market & competitor intelligence team).
Working with a variety of clients, I more rarely see a focus on really explaining the financial drivers of business to analytics teams. Surprisingly, given their high numeracy skills, businesses tend to keep that education & information within their Finance or Strategy teams. This is well worth addressing. Investing in commercial training, sometimes ‘in house’ from strategy & finance leaders, can pay huge advantages in quality of future analysis. If you need to do this on a ‘shoe string‘, then I’d also recommend “Naked Finance” by David Meckin as a great book on this topic.
In line with the benefits of collaborating with Finance & Strategy teams, it can also be worth asking your Sales leaders to help. Any effective Sales manager is well used to keeping his team up-to-date with latest sales performance, or expecting account managers to ‘know their numbers‘. A similar requirement, to understand headline financials & be able to quote the latest KPIs that matter (often sales performance), is a great expectation for analytics leaders. I have found it really helpful to expect insight business partners or those managers with responsibility for a particular business product/channel/function, to ‘know their numbers’ and be able to talk meaningfully with their commercial contacts at any time.
Doing the rounds, using team meetings to achieve more
When asked about the aspect of their job that they like least, many an analyst (or leader) will cite excessive emails or meetings. Keeping the latter interesting, is a key challenge for any manager or leader. When meetings work well they can be a really effective way of keeping a whole team up-to-date, on key knowledge they need, and together spotting opportunities for collaboration, challenges or new ideas.
But, in line with the entrenched functional fiefdoms of corporate life, too many team meetings are too insular. They may go an entire year or more without anyone from outside that team or function attending. This is a missed opportunity, as encouraging leaders to ‘do the rounds‘ (by visiting other team meetings) can be an opportunity for a win:win result.
Relevant to our topic of increasing domain knowledge, a simple solution can be to ensure that your analysts have opportunity to hear first hand from leaders or experts in other business functions. The quickest way to understand sales performance, marketing strategy, customer irritants at service touch-points or potential of IT systems – is to hear directly from the function with ownership for that aspect of your business. The skill here is to create a mutual ‘safe space‘ sharing environment, where other functions leaders or experts feel valued, and your analysts feel unembarrassed to ask basic questions about anything they don’t really understand.
I mentioned a win:win outcome. The other benefit that can come from this approach is to give your team a break from you. I say that because I recommend reciprocal visits. This gives opportunity to, at the same time, demonstrate valuing of other business function & create an opportunity to talk directly to their direct reports about Analytics & any key insight you need to champion. It can also really aid the development of your team to be exposed to a variety of leaders and/or for one of your team to have responsibility for running team meetings in your absence.
So, there is another example of something the leader can do to create environment for analysts to build domain knowledge. But with all these 3 initiatives, there is also a responsibility on the analysts to engage and learn. Solutions should not be ‘spoon fed’ too much, as everyone is primarily responsible for their own self-development. That emphasis comes to the fore in my last suggestion.
Getting out & about, domain knowledge & real customers
The final means of encouraging analysts to gain domain knowledge, is to expect them to meet customers on a regular basis. Something I’ve seen work really effectively is to expect every member of your Analytics team to hear/meet customers at least once a quarter (and include that as a metric in their performance assessment).
Human contact is always a more emotional & immersive experience than reading research reports or knowing the results of behavioural analysis. For that reason, although listening to calls from a call centre, or viewing focus groups in a research studio, can help – the best solution here is meeting customers in their natural environment.
A few insight teams may have the advantage of occasional ethnographic research, if so, I encourage analysts to also get involved (if possible). But for most sectors, a more realistic goal is for analysts to visit the key distribution channels & observe or talk to customers. For your business, that may include stores/branches, call centres, advisor/intermediaries (sometimes visiting customer homes), mobile units or partners who supply services.
At a minimum, analysts should experience using the mobile and web applications themselves (for sales & service experience). But if it is possible for analysts to engage with typical customers in the normal environment for them to buy or use your products/services, it can create opportunity for emotional insights. In a previous series, focussed on insight generation for proposition development, I mentioned how this type of knowledge can be very useful to converge with internal analytics or research studies.
Beyond learning further domain knowledge, often about the real world context of data items or customer behaviour patterns observed in the data, such visits can also increase motivation. To truly empower your team includes creating opportunities for them to discover their own “why“. Why do they choose to keep working for your business? Why do they do the work they do? Seeing the difference it can make to customers lives is often a key to unlocking such personal motivation too.
How do you develop domain knowledge in your teams?
I hope the above content as convinced, ore reminded you, of the importance of domain knowledge.
How do you currently ensure your analysts or researchers gain this knowledge? I’m sure there are plenty of good & better ideas out there.
Please do share your thoughts and suggestions in our comments section below.