What are the questions, that insight leaders today, want answered? (part 1)

Yesterday, I had the privilege of chairing a table at a DataIQ Discussion event. It was a great opportunity to hear the questions that insight leaders today most want to be answered.

Not surprisingly, I was invited to chair a table on the topic of ‘Customer Insight‘. There were about 70-80 data & insight leaders attending this event, each having 8 tables to choose from, for their 3 ’roundtable sessions’. So, it was pleasing to see that 21 leaders chose to join the conversations at my table.

I’m grateful to all who did so, as we had some really interesting & passionate debates; on topics that mattered to them.

So, for any readers who did not manage to get along to this engaging event, I thought I’d share both the questions & answers those leaders offered. As they came from a wide range of sectors and diverse data or insight roles, I hope there is something here for everyone.

It was such a useful event, that I’m glad to say we have enough content for 2 blog posts. So, here is part 1, of the questions insight leaders are asking today (with some ‘wisdom of the crowd‘ initial answers)…

How do you measure the impact of Customer Insight (to demonstrate value)?

It was widely agreed that this was easier to measure for database marketing applications (like building predictive targeting models), where marketing effectiveness measurement can give you an ROI. Where it was harder, all agreed, was measuring the impact of more strategic insight work (to guide strategy or propositions). Interestingly, no mention was made of econometrics, so perhaps that is still not widely used.

A few tips, from leaders with this challenge, included:

• Understand what matters most to your business right now, focus on that (as we’ve said before for influencing top table)
• Think about who you need to sell any metric to (choose your sponsors, to guide measurement design – what do they measure?)
• Combine ROI or hard financial metrics with compelling customer stories (bringing benefits to life for C-Suite thorough research as well as analytics)
• Tell stories that matter to your Exec (in their language)
• A helpful alliteration can be: Strategy > Story > Stakeholders > So-What?

How do you break through cultural barriers to drive action?

Being dismissed by those who have been there longer than you, was a common theme. The barrier that long-tenure & groupthink can cause a “because we’ve always done it that way” mindset, that is resistant to change or data evidence. But it was also acknowledged that insight leaders also had to pick their battles, some of these skeptics could be left to continue for now. Know what you can influence & make a difference there.

A couple of leaders raised the fact that it was, nowadays, more affordable & viable to just pilot ideas yourself. “Stick a PC under your desk & get on with it“, was the advice from those who has succeeded with such ‘under the radar‘ approaches – until they could demonstrate value. That approach has certainly worked for me in the past.

There was some skepticism that the much-lauded Next Best Action decisioning capability, with those newer to insight execution seeing this as taking too long to build. Instead, the most popular focus was on identifying moments of truth in customer journeys to fix. Once again, there was a reference to capturing the emotion expressed by customers, to help make the business case for acting on insights into ‘irritants’. Joining up research & analytics offers the potential to empathise with why it matters to customers & track changes in behaviour once service experience is fixed.

A final barrier, that was identified, was the need to develop analysts beyond being ‘order takers‘. More than one leader cited the need to improve questioning skills, so that analysts got to the ‘real business need‘ that often hides behind the data or insight being requested.

What is the best Organisational Design for effective customer insight team?

Another perennial topic for insight leaders & one that has previously features in our posts & polls.

At this event, there were some advocates of an ‘Insight Centre of Excellence’, because of these perceived benefits:

• Consistent guidance
• Economies of scale
• Technical development, easier collaboration & coordination
• But advised use of ‘Business Partners‘ as well (or a ‘matrix‘ link into business teams)

This time there were less fans of a federated model, but other benefits identified for it:

• Maintaining domain knowledge (local business/products/channels/issues)
• Easier for analysts to see the difference they make (motivation/engagement)
• A few voices for a combination of both solutions, especially globally

I must say, those answers surprised me, as I’ve seen far more clients see benefit from federated insight teams. But, it is ‘horses for courses’, especially when considering size of team and organisational maturity in using insights.

Which tools do you use to generate customer insight?

The answers to this question offer an interesting update to our (now rather dated) poll on analytics software used.

Before getting into specifics, we agreed that the categories of technology solutions needed were:

1. Somewhere to store all that customer data (Data Storage solution)
2. A means of cleaning-up data, moving & merging (Extract Transform & Load, ETL solution)
3. A way to access this prepared data, for exploratory data analysis (Data Access solution)
4. Specialist software or programming language for statistics & analytics (Analytics Software solution)
5. Means of turning numbers into compelling visuals for presentations (Data Visualisation solution)
6. How the insights/models/triggers discovered can be deployed (Execution solution)

I think they covered most of the bases there (at least from business perspective, rather than ‘technology stack’ thinking in IT teams).

On that basis, here are the tools they referenced as being used to meet those 6 technology needs:

1. Data Storage = SQL Server, Netezza, SAP, Informatica, DB2 or Amazon Redshift
2. ETL = SAS Data Flux, Talend, Data Stage, MDS or ‘in-house’ programming
3. Data Access = SQL client for database solution or Big Query
4. Analytics = SAS, R or Python
5. Data Visualisation = Tableau, GG Plot Clickview or still just Excel
6. Execution = Adobe suites or various other Marketing Automation solutions

How have you executed on your insights? Driven action as a result?

This should always be a key concern for customer insight leaders. We’ve shared before some thoughts on maintaining a ‘culture of action’ in your teams. So, it was interesting to learn what had worked for these other insight leaders.

They had diverse experience, tailored to the needs of the customers & businesses they served. But here are some of their real life examples:

• Built a Churn Model (reduced churn by 15%!)
• Defined Insight so the whole business had a common definition/language ( had to be Credible/Novel/Actionable), that set exec expectations too
• Bridge the gaps between silos (acting as the glue bringing other teams together) to enable execution of insights
• Work closely with internal customers (e.g. getting insights embedded into comms design)
• Choose several smaller problems & get results out quickly to prove value (incremental gains)
• Not just a matter of deploying a model, but being able to explain ‘why’ customers act that way
• Use learning from research to inform where to use analytics (converge evidence)

Insight Leaders today, did they miss your key questions?

I hope sharing that Q&A was useful & relevant to the challenges you face in your role.

Without giving away the content of Part 2, there are questions still to come on definitions, best practice, techniques & making your analysts’ life easier!

Have we missed anything that matters to you? Do you have a burning question, as an Insight Leader, that you’d like to share or debate?

Feel free to comment below and we’ll be happy to get the conversation started.