Don’t allow your insight to be locked down & isolated from your marketing

Continuing our theme of choosing your technology stack, here is a useful reminder to not allow your insight to be locked down.

By that term, I mean the insights and models produced by analytics or data science teams. Often these are intended to improve the performance of marketing communications or improve the appearance of personalisation.

In this post, Marketing Operations expert Simon Daniels reminds us that these need to be delivered to Marketing in a way that works. Otherwise good quality technical works risks being locked down & isolated from use elsewhere in your business.

Simon is a Marketing Operations consultant, speaker and writer. He’s shared the Marketers perspective with us before, on topics including working from home, shiny new tech & digital collaboration tools. Now over to Simon to share how to not allow your insight to be locked down…

The workplace challenge we face

The pandemic crisis has seen us all confined to our home offices, kitchen tables or even sofas. Working away in isolation from our colleagues in the interest of suppressing transmission of the virus. Fortunately, modern technology has meant of course that in fact, we have avoided complete isolation. We’ve been able to continue sharing, collaborating and exchanging ideas, plans and actions.

Insight should be treated in the same way, whether in extraordinary times or business as usual (whatever that means!).

That technology stack poll and results

There’s been some great advice so far in this Customer Insight Leader series on choosing analytics technology. I thought I would contribute some thoughts from a marketing perspective, and in particular marketing operations, which is my area of interest.

Clearly, insight can be applied right across a business, from proposition to customer experience, logistics and more. As such, much of what I’m suggesting here is doubtless universally applicable.

In the interim results of the data leaders’ technology stack poll, it’s little surprise that data analytics has emerged as the leading technology component. Data visualisation also came out as being an important aspect to insight leaders and I suspect were the same poll put specifically to marketing leaders the results would be the same. Certainly, the heads of marketing I work with are always concerned with measuring marketing performance and being able to readily digest it.

Don’t overlook this often missing part of your analytics tech stack

The aspect that is easy for insight specialists to overlook though, and this applies beyond marketing, is how the output of analysis will be operationalised. Viewed from the perspective of B2B marketing operations, where I’m mainly focused, the world tends to revolve around a marketing automation platform (MAP). This is closely aligned with a CRM system of some kind.

The MAP is where outbound campaign execution takes place. Responses are captured and certain other digital activity coordinated. Any insight, segmentation or treatments arising from analytics, modelling and other analysis needs to be easily integrated with these tools. This means that an efficient, operational mechanism must be put in place. One that returns insight to these production systems to support campaign activity.

No more text file extraction for Marketers

In particular, any kind of reliance on text file extraction and transfer is completely unacceptable. This represents far too onerous a process to be undertaken on a regular basis. That’s regardless of ostensibly how straight forward it may be.

Front line marketers are especially unlikely to be impressed at the suggestion of having to work this way. Marketing operations teams are already too hard-pressed to be handling such manual activities. In addition, an extract of a particular cohort of customers to be included in a particular communications programme, will likely already exist in the marketing system. The task is to flag those individuals, not merely load them into a database.

What’s actually needed is to be able to systematically flag selected cohorts in the MAP or CRM system, for inclusion in any new or follow-up campaign activity. This should also, if at all possible, avoid the creation of endless flags. Ones that are only populated in a tiny proportion of records in the overall database. Preferable instead is to make use of whatever list functionality the system already has, to represent a selection of individuals in a campaign. 

Save the workload with APIs

It should be possible to achieve this through the ubiquitous APIs that all modern platforms provide so that the data interchange can be entirely automatic and seamless. (Scott Brinker, in his guise as Chief Marketing Technology Officer, reported on some interesting research on cloud-platform integration recently.) 

Having been created in the analytics platform, a segment or audience selection can be pushed back to the marketing system for execution. In this way campaigns can be built in the marketing system in the normal way without end-users even necessarily having to have a particular understanding for where the insight originated. Even automated campaigns could make use of such insight as part of recurring or on-going programmes.

Remember deployment in your selection of analytics tech stack

Returning to the point of selecting analytics technology. This represents a key requirement that must be supported by your chosen solution. When shortlisting solutions and providers, make sure that the operationalisation of insight is front and centre. It needs to be in your requirements definition and clear how this will work.

Have analytics/insight solution providers demonstrate how their platforms and tools will integrate with your MarTech stack. Involve marketing operations or whoever is responsible for the stack so they can confirm it performs as promised.

While such a proof of concept is vital, heed the advice of marketing technology analyst and all-round guru David Raab who cautions against misusing the POC in system selection. Insight created in splendid isolation and subjected to digital lockdown is no insight at all. Make sure it can find its way back to the wider technology stack where it can be put to use in the real world.

How are you ensuring your insight can be deployed by Marketing?

Many thanks to Simon for this important reminder. I’m conscious that a number of data leaders whom I’ve interviewed on the podcast have emphasised the need to focus on execution. It has also often been a key theme in conferences.

For a data leaders perspective on this challenge, it is worth also checking out the practical advice in:

• “Guerrilla Analytics” by Enda Ridge
• “Practical DataOps” by Harvinder Atwal

What about you? Are you confident that your insight is not locked down? Are you integrated with MarTech systems through APIs or other solutions to avoid manual workarounds? If you have advice to offer on this aspect of the data leaders’ technology stack, please let me know.