Want to make sure your analytics implementations go well? Does 66% user adoption sound like a dream? Put into practice these 14 best practices and tips to cultivate user adoption around the data and analytic systems in your company.
It is always surprising to hear about a company whose first BI deployment failed completely.
They didn’t fail because of missed requirements or underwhelming business impact, but rather a full-fledged flop, with ZERO consistent users of the initial app. How does this happen?
Here sits an Analytics Director, or equivalent, having pledged the sun, moon, and guided analytics to Senior Management, but with nothing to show except the ashes of what used to be thousands of dollars.
While complete failures tend to be outliers, it does happen.
So how can you avoid calamity? And how can you best optimize traction of your analytic deployments?
A Director of Analytics at a pharmaceutical company shared some of her strategies for fostering user adoption when her team rolled out their new BI self-service software after their old system became too bulky and slow for effective use.
She believes that their early success came from approaching the implementation as a complex change management issue, worthy of a formal strategy, even during the experimental phases. Today, they have 600 total users: 200 external users and about 400 internal users, approximately 2/3 of the company… and growing.
She shared some advice and best practices to cultivate user adoption for your BI systems throughout your company:
Be very choosey about who the initial audience will be. This company decided to focus on Business Development & Client Services first because they were the most frequent requesters of data, not necessarily the largest or most complex.
The business users know best for what data they need, how often they’ll need it and how to output it to their workflow. However, there are often hidden nuances to a set of requirements, so don’t just take them as a task list and start developing right away. Some caveats, if unaccounted for in the initial app, will drastically decrease value and discourage adoption.
Deliberately target the data consumers themselves, not only their department head. The people in the trenches know the need more precisely, and their input is important for tailoring content and usability to fit their workflow. Additionally, as they invest their own time and interest, data customers are more likely to engage with the finished or even preliminary product. As one of Analytics8’s account executives likes to say, “make them feel part of the solution.”
Don’t simply engage users during requirement gathering, and then drop an app in their lap several weeks later. Keep them updated on development. Their data needs are often recurring, and some very useful input can come up later when progressing through a business cycle. Try to think of their input as scope improvement rather than scope creep – as unusual as that may sound.
Initial apps should include extremely targeted and rudimentary reporting objects. Try meeting a high volume and repeated need, which entails more frequent visits to the app, boosting familiarity. More complicated, but lower volume data questions naturally follow onto the initial development, as employees realize the full implications of the technology.
You should build the apps keeping in mind: “what will require the least amount of training?” The fewer clicks needed, the better. More elaborate UI enhancements come later, once a baseline knowledge and comfort is established with the new tool.
The temptation is often to leap immediately into visual analytics, which is how software is sold to management. However it’s important to realize that most users, at least in the case for this pharmaceutical company, export numbers after a few quick selections, and shuffle off a report to someone else. Quick and intuitive are primary goals – train well on export and sharing methods.
There will be excitement around the initial release … and that’s good. However, end users inevitably find discrepancies between the new self-service software and their old method of reporting. This can leave them skeptical and turned-off to use the new software.
Be prepared to address how old numbers are wrong and why. Make sure you do this in a timely manner and communicate it widely and clearly.
Make sure to open up lines of communication with well-advertised opportunities for education and training. This company went to the extent of adding unique branding around their internal training program, making it stand out and be special and then invited employees to attend.
In these sessions, make sure your coworkers understand what they are capable of doing versus what they’re not capable of doing.
Advertise availability and encourage attendance for one-on-one training. For example, you can have an open door policy for office hours every Friday from 1-3pm. This is where power users will come out of the woodwork, some unexpected. Make sure everyone has this opportunity to not only learn, but to discover their own capabilities.
Create a forum, time, and place for internal users to share their experiences, realizations, ideas, and questions. You can even add an element of gamification (i.e. award Girl Scout cookies for the highest usage). “Just make it a little fun to keep people engaged,” said the Analytics Director.
Replacing an old system is not a scenario where we can just “rip off the Band-Aid.” But as you’ve put these other support avenues in place, you can begin to decrease your attention to the old system. Refresh data less often and then eventually shut off the old system, while keeping the apps available for a grace period.
REMEMBER: Some people will be reluctant to change. You eventually need to start shutting off old systems and reports. Put a pull and push system in place by making the new software’s app content available sheet by sheet while you simultaneously turn off the similar report in your old system.
Various departments have different data consumption “personalities” and need to be served differently. Finance, for example, is a “creatures of habit” type of department. They prefer reliability to amenities in their data applications, and are more willing than others to undertake manual pains to stay with an old system whose numbers they respect. In some cases, they will default to the old platform until it is simply no longer an option.
When creating the “source of truth” for your company, it is important to arrive at a consensus of what actually constitutes the truth. For example, a Sales App designed for use across all departments might suffer adoption issues when there are variations in how they each calculate and consider sales. Make sure to have a regular meeting with SMEs from various departments to keep your data reliable and trustworthy.
#UserAdoption Success Tip: Approach #implementation as a complex change management issue, worthy of formal strategy http://ow.ly/ch6V306JB3d
Even though self-service BI software’s boast is “rapid time to value” and “easy and intuitive interface,” this can be misleading and a real challenge to the average corporate data consumer if the software is not properly introduced. Be sure to manage your company’s expectations and clearly communicate the obstacles and strategies for overcoming those obstacles by putting these practices in play to encourage buy-in from your colleagues, ensuring their success and your return on investment.
To thrive with your data, your people, processes, and technology must all be data-focused. This may sound daunting, but we can help you get there. Sign up to meet with one of our analytics experts who will review your data struggles and help map out steps to achieve data-driven decision making.