Digital Analytics Hub Conference – Part II

In my first post about my experience at the Digital Analytics Hub Conference, we covered the event’s unique discussion format of “conversations, not lectures.”

The overall theme of stimulating thinking and encouraging an open, honest dialogue among attendees was one of my favorite aspects of the conference. For those that were unable to attend, below are key takeaways from a few sessions I attended:

Experimentation on Steroids

Experimentation/Testing programs have been operating within many companies for nearly a decade now. That is a sufficiently long enough period to exhaust most, if not all, of the “low hanging fruit” of testing and targeting potential. There are new opportunities with the introduction of new technologies such as mobile, apps, and an increasing variety of wearables. However, in most cases the fundamentals remain the same.

So what’s next for experimentation? Where is the next growth area? In this huddle, we explored the opportunities for getting better results, driving more informed decision making, enhancing customer experiences, and avoiding common mistakes. This conversation focused on practical ideas and concepts aimed at improving our experimentation programs to ensure continuous success.

Takeaways:

  • If you’re not testing, you’re behind your competitors – BUT – you must INVEST the capital (people, time, $$) knowing it might not immediately drive results
  • Testing is about driving results, but ALSO about challenging policy/norms
  • When testing, nothing should be off-limits – companies often have foundational things they don’t want to test against
  • Even when an “old” test wins, it is often valuable to keep it running at a small % – this can help with tying cost to testing
  • Testing/experimenting is a much better sunk cost than rebuilding the site/marketing on a poor site or concept

 

Setting Effective Targets

Setting challenging, but attainable targets, is proven to improve productivity and create a high performance culture. Targets foster engagement and encourage focus on the most important tasks. Furthermore, true data-driven decision-making necessitates clearly-defined targets – you cannot determine the best way to get from point A to point B if you have not yet defined that point B, AKA the target state.

In this huddle we discussed techniques for setting effective targets for your KPIs. Topics included:

  • Which KPI to consider
  • Setting targets that align to broader business objectives
  • Discussing the difference between targets and forecasts
  • How understanding the difference between targets to forecasts can be used to drive action
  • Coordinating action planning with target setting

Takeaways:

  • KPI targets should contain a change %, and a timeframe (both for comparison/baseline as well as how long to measure)
  • Measurement down to the tactical level is something many strive for, yet often fail to achieve because of competing priorities of the business or client
  • Effective targets are “local” – a goal for a tactician that they have no control over makes moving the needle challenging
  • If you company or agency has multiple “attribution” models, one (and only one) needs to be agreed upon – single source of truth allows for best measurement and movement of business goals

 

Where the Girls Are

I think this topic deserves an entire blog post (or series) by itself and is something that you’ll likely see come out of Delphic in the months to come. Gender diversity in analytics and technology (or in this country in general) is a touchy subject to some, but it is paramount that it be discussed at length. As a father of a daughter and a father of a son, I think this topic is something to discuss within every organization, within every family, and is very much needed.

Last fall, a popular analytics podcast took on the subject of women in analytics. The lively discussion had me arguing aloud in my car with the panelists, and rattling off a list of prominent women in the industry.

Let’s continue the discussion in person. Is there really a lack of women in analytics organizations? Or is that a myth? The data indicates nearly half of all analytics professionals are women. Perhaps it is a marketing issue – how do we help women build their personal brands? What can we do to mentor our younger analysts to ensure we build the next generation of analytics leaders? How do we attract and retain more women to our field and our management ranks?

Takeaways:

  • Women are NOT underrepresented in analytics but there IS a reason why people seem to think so (see next point)
  • Women are far less prominent in the speaking circuit and in the blogosphere than many of their male counterparts – see this recent Forbes article
  • Read the story about “amplification” practiced by women in Obama’s White House to make sure ideas are heard no matter the person
  • It is crucial to know that your voice is valuable and to never let anyone treat you otherwise
  • Vital to have the tough conversations within your organization – failing to recognize there might be a problem is part of the problem – hint: you likely have a problem…
  • Flexible work solutions could be an extremely helpful way to grow the number of female applicants and employees at your organization

 

Return on Investment in Analytics (RoA)

Competition in the digital world is intensifying as barriers to entry keep dropping. With the advent of digital-savvy management, the desire to consume data is increasing. Data is quickly becoming the next battleground. Data analysts along with data scientists are the most sought after soldiers. So are we delivering the expected return on investment in analytics, or RoA?

Organizations are constantly challenging their teams with this exact question – looking to tackle the challenges of prioritisation and selection of analytics initiatives. The measurement of RoA, as contribution versus cost, is not an easy one. Digital analytics teams commonly serve multiple stakeholders and depend on them for any recommendations’ execution.

In this huddle we worked with examples and experience from attendees to explore the methods and challenges for an RoA calculation. Topics included:

  • How and whether it is necessary to quantify the distinct contribution of the analytics team
  • A look at the investment and organizational processes required to enable any RoA
  • A review of the desired outcomes and how they would vary between different types of organizations
  • Why knowing your RoA can further leverage the strategical position of digital analytics

Takeaways:

  • Run tests for tests, not full business results – it can be dangerous to extrapolate out to an entire business or line of business
  • Work with clients (internal or external) on a survey specific to analytics – value, need, etc.
  • The cost of purchasing third party panel data far outweighs the internal/agency business cost of analytics doing analyses
  • Not just comparing benchmarks to one’s own numbers, but industry as well
  • What can your analytics team do that would otherwise be purchased or outsourced?
  • Data collection, normalization, storage, governance, and attribution are part of company equity

 

I hope that you’ve found some value in what I’ve shared. As the leader of our analytics practice at Delphic Digital, this conference confirmed for me that many of the things we do for our clients are far advanced versus what other agencies are doing. It also illuminated for me areas of improvement that we can implement to ensure that we are driving business results even more than we already do.

I am always up for a conversation on any of the topics covered at the conference (huddle schedule here), so please reach out and let’s talk analytics!

  • Ken Killian

    Great read Andrew!
    -Ken

« Prev Article
Next Article »