The time is right for Marketing, IT & BI teams to move towards closer collaboration, to gain an increasingly joint ownership of the customer data, exploiting each other’s strengths in engineering, strategy, governance and insight to benefit the company as a whole.
We’re loving the many articles/posts being written currently by other consulting firms listing their definitions or conditions for ‘agile marketing’, and we agree with much of what is being said about the methods for achieving an agile state within a company.
But it’s worth reflecting on how the marketing and BI teams have arrived where they are today, to recognise this natural evolution in company data science. Technologies for understanding audience data have become ubiquitous and commoditised.
Previously, companies have engaged in ‘big IT’ projects, such as a move to a ‘single customer view’ database, believing they can accelerate a company’s progress.
We would argue this is simply approaching new ways of working with old thinking… It’s far more productive for many companies to ‘go ugly’ – that is, hack together insights from the available data and make iterative progress without waiting for grand software projects to deliver the promised elixir.
That’s not as unprofessional as it sounds. Companies are increasingly making provision for the data to be accessible and today’s ETL tools enable data-hacking with for most inexperienced of operators to mine insight from their data warehouses. So, why is ‘agile marketing’ just beginning to be spoken about? We think we’re seeing a move from the silos to cross-department collaborative working teams; the best-practice for which is an emerging approach which requires a ‘scrum master’ to bring the disciplines together and convene an agile method.
While the technology is important, it’s the behaviour and coordination that makes the difference. A well co-ordinated cross-discipline team has the opportunity to leverage the data and technology much more effectively, as with software development where agile methods are commonplace.
We expect to see the rise of the scrum-master specialising in leadership of data-driven marketing teams from across the company. To achieve the successful iterative approach to incremental gains across the customer profiles, to set a culture of respectful idea sharing, the scrum-master needs to be experienced at preparing, convening and driving the scrum meetings which co-ordinate the collaboration. With such a leader in place, only then, do we encourage our clients to ensure they have all the other ingredients for agile marketing.
To create the right conditions for Agile Marketing, we have found these ingredients vital:
Be absolutely clear about what business-critical problems you are trying to solve or the data-driven hypothesis you wish to prove or disprove in order to create value that the business can test. We recently met a prospective client who wanted to establish a ‘data business’ within their publishing house from the ‘data lake’ they have already assembled. We have proposed a workshop for key stakeholders (including their subscribers) to first establish where the value is, because we can only hope at this stage, that the data lake contains data points that can create value.
KPIs in stone
C-Suite can make this easy: define a core small set of KPIs that the business is optimising itself on. We have spoken to clients from both sides of this story recently: a retail bank which has ‘turned itself around’ because the new CEO set just a dozen KPIs for the whole business. No investment was available unless the business case and planning scenarios could be shown to predict uptick in one or more the KPIs. Conversely, a major global CPG manufacturer, without a strong mandate for local markets to focus on local optimisation against a standard set of KPIs has found itself with an array of data-points with no structure for comparing different market nuances or media initiatives – a major struggle to introduce before the agile method can be used to optimise.
It works so much better when led from the top.
Test & learn culture
Only two years ago, we worked with a media agency that never reflected on the campaigns it supplied our client. If you don’t test variables, you don’t learn anything. This has to be a fundamental of the team’s ethos.
Understand and acknowledge that agile marketing programs always have a follow-up question after the sprints, they rarely ‘finish’ per se, but they do provide iterative optimisations for on-going customer acquisition and retention channels.
List & prioritise questions
In order to make sure the right questions are being worked upon, the hypotheses or challenges of the most value to the business, first a list of questions needs aggregating. Then they need prioritising. All stakeholders, junior and senior in-house or agency-side, need to be able to contribute questions and hypotheses for consideration by the team. To give visibility, priority and encourage the collaboration, we wrote our own tool; TestBoard, to drive productivity.
Some organisations treat analysts as a luxury. They never have enough. This isn’t surprising given their spiralling market rates. A solution we’ve seen executed well is the building of a freelance team, to investigate specific problems and be disbanded at the end of a sprint. It’s a productive and motivating way to push analyses of specific problems.
Always-on data warehouse & models
Even econometric models of the twice-annual variety are being migrated to always-on infrastructure, so that individual channel and market tests can be conducted in more frequent timeframes to achieve incremental gains.
We’ve built a number of dashboard integrations on a variety of BI and entry-level tools for clients. There are pros and cons about all the offerings, but the most useful thing is to have the full set of KPIs reported on a weekly basis, available on all devices. A special mention for the Domo product launched this year to a wider market. This looks like a ‘must have’ solution.
Weekly scrum, scrum-master
Quite simply, it is a ‘crime’ within a successful agile team, to turn up to the weekly scrum without having updated the scrum board, knowing the metrics changes for which you are responsible and having considered the KPI dashboard on the bus on the way to work. If you are not informed, you start the meeting catching-up, this is frustrating for those who have prepared for the meeting.
By introducing these ingredients, and driving collaboration between the vital functions, cadence increases, and so do gains for the business.