How can you drive Tableau adoption? 4 red flags to ensure success
You've just introduced a new piece of analytics technology to the company. The proof of concept was extraordinary. You saw things in your data that would have taken weeks in Excel. Analysts are trained up, departmental heads are briefed and IT gives the nod you're good to go. Time to roll out across the company. The data crunchers of yesterday are about to become analytical and business heroes. Six months later, things look a whole lot different. Usage of the technology is declining, projects are stuck once again with IT and apathy is spreading. So often we see optimism turns to frustration when investments and the promise of technology aren't realised.
Last week at the 2016 Tableau's partner summit, I had the pleasure of listening to Vikram Ekambaram, Strategic Account Manager in the Global System Integrator team at Tableau talked about Tableau's Drive methodology. This methodology is aimed at driving agile analytics projects, relying on a fast iterative approach rather than traditional long-cycle deployment.
If you're about to start or are in the middle of your analytics project or technology implementation, here are 4 red flags to make you stop and think before you charge ahead:
1. "We have to finish our data warehouse before we do any analytics"
You just bought a data visualisation software. You know the visualisations and insights it will provide are only as good as the data underneath. So you dive headfirst, investing everything into your data warehouse project. Fast forward to the end of your data project. Business users start accessing the data. Suddenly you find everyone is using your data warehouse in ways which are completely unexpected and irrational. Frustration spreads as you find out even with your new analytics software, you aren't able to address the latest business needs.
Using data well does not mean an all or nothing approach. Start small. Prioritise different areas of investigation. Ask your end business users the top 10 questions they want to answer. Based on these questions, select areas of data and build dashboards as you go.
Light-touch self-service data blending tools like Alteryx and Trifacta are now allowing you to independently access, cleanse, and blend datasets using drag-and-drop functionality. This means you can rapidly prototype and test your solution with your end users as you go, instead of waiting for large amounts of development work. The process will not only reveal how users will interact with data, but will also surface unknown data quality issues, therefore reducing risk and speeding up time to value.
2. "We have self-service analytics we don't need IT"
One unique selling points of self-service analytics technologies is the ability to bypass IT. This is great, and true for small projects. You can go get your data, plug and play. The moment you try and expand to bigger, enterprise projects this approach can come unstuck.
Bigger projects means increased data governance, security and scale. Without IT, big projects are at risk of getting stalled due to shortfalls in addressing these issues. So at the start of your project, give IT ownership to build a data model to support your dashboard requirements. This means when you scale out, dashboards across the business can point to the right pieces of data within the model, rather than requiring lots of different access points to different portions of data which can result in multiple versions of the truth. The role of IT today has increasingly shifted to data governance.
3. "We are utilising an agile approach once the business sends us the requirements"
Agile analytics methodology is borne out of agile software approaches.This does not mean having a disconnect between the business and the technical process. If business processes are gathered and then left for the analyst to interpret in a black box, the result is exactly the same as the one agile methodology is trying to avoid in the first place– a solution which doesn't suit the end-users requirements.
This is why analysts or technical experts must collaborate thoroughly with executives when using the agile approach. They need to have a clear discussion about what data the executive wants to see and how it is best analysed in practice with the selected technology. By building the prototype dashboard together, you can drag and drop data across the screen and explore different views to discover where the 'magic moments' lie. Having agreed on a prototype, the technical team can then go away to tighten and improve on a solution that actually brings value to the business.
At Concentra, we developed our own agile project delivery method called "ProAgile". The method is mostly based on the principles, ceremonies and artefacts of SCRUM but combines elements from the Project Management Body of Knowledge (PMBOK). As you might imagine the extent to which "ProAgile" is well defined, understood and adopted by all involved is key to the success of any project. This is why we use the LEGO Game (see Figure 1) as a way to spread knowledge and familiriase people in the company with the concepts of the agile techniques.
4. "I can roll out self-service analytics without a data strategy"
The funny thing about analytics is that you think it will answer all your questions. The irony is that the more answers you get from your data the more questions you will want to ask. In a business scenario, this typically results in having to go back to the IT queueing system for more data.
Without a data strategy you will only see success in the short term. At the start of your project, think through your data roadmap - how can you get more access to data and increase the scope of your project? Once again, this means involving IT from the start, not simply bypassing them. Without a data strategy your self-service project will stagnate.
Setting the right focus
As we saw in Gartner's BI magic quadrant this year, the market has changed. Self-service analytics is now the name of the game. As technologies increasingly shift towards this model, the question will turn from simply "how can I understand my data?" to questions of "how can I get more value from my data?" and "how can I get everyone across the business getting value from data?" Those that will be likely to succeed will not be the ones with the biggest budgets, but ones with the ability to drive user adoption and impart real value to the business through their analytics projects.
Download your free trial of Tableau and discover how you can drive better analytics insight from your data.