The Data Café - A Secret Sauce for Adoption
A STRATEGY TO DRIVE USER ENGAGEMENT
…and promote end-user adoption of data solutions once they’re deployed & live
OUR SUCCESSFUL DATA SOLUTION
When do we know that we’ve created a successful report?
Is it when the report is performant & fast? What about when it looks nice, because it follows best practices? When it’s accurate and correct? How about when it’s delivering us predictive or even prescriptive analytics?
We might be thinking all or a combination of the above. That might be true if we are defining a good quality solution, but is it a successful solution? Ultimately, a data solution’s success is defined by one thing: is it being used to derive business value?
IF NO ONE’S USING OUR SOLUTION…
…THEN IT’S NOT A GOOD SOLUTION.
Our solutions can be all of the things in the first paragraph — pretty, performant & powerful — but if they’re not being used, they’re bringing no value. Usage is arguably one of the bottom-line definitions of success. It’s why the usage metrics of our reports are so confronting. So what can can we do to increase usage? There are many factors involved, but we might look at three different things:
Design the right solution for the business questions addressed;
Collect the correct requirements from the correct people who will actually use it.Build a high-quality solution that fits user needs and expectations;
Follow best practices to create something efficient, accurate, innovative & accessible.Create & cultivate a data culture around the solution, supporting it & the end-users who use it
In this article we will look at #3 - an using “Data Cafés” to cultivate this data culture & improve adoption.
WHAT DO WE MEAN BY ‘ADOPTION’ AND WHY DO WE WANT IT?
Adoption is a complex and nuanced topic. In a nutshell, your solution - your report - is ‘adopted’ not only when the end-user community is using it. Adoption is more than that. Adoption implies an acceptance or integration of the solution into that community’s process and decision-making ecosystem. Healthy adoption also spreads; in the best-case scenario, it’s not just accepted, but advocated by end-users to others.
We like to think of data solutions as these mechanical tools. We pick them up when we need them, we put them down when we don’t. We use them. We walk away from them… we replace them, we throw them away. We often imagine tools, gears, machines… cold, lifeless metal. But is that the best way to think of them?
In reality, data solutions are organs in the living body of our user community; our organization. They are tightly coupled with our processes, our way-of-working, and even our identities in the workplace.
These data solutions keep the body healthy when they function well, and cause pain and suffering when they don’t function well. Infused with data lifeblood, they keep our organization alive. This is why introducing a new data solution into a complex decision-making environment feels like surgery, and is sometimes such a delicate, painful process. We are performing a transplant; removing one organ and replacing it with another that we created. There’s a fragility there, not only in the surgery but also afterwards - will the organ be accepted by the host?
Obviously we can ensure the surgery’s success by preparing properly, getting the right requirements & doing a good job. But like any transplantation, much of the work comes afterward; aftercare & recovery. We can prevent infection (i.e. bugs & bad quality data), or conduct small interventions where things aren’t going right (i.e. change requests). But the best way to ensure recovery is to look after the patient, themselves. To talk with them, follow-up with them, rehabilitate them, consult them. It’s a process of care and empathy between professionals and end-users to help them function normally - or even better than normal - compared to before the surgery. Doing so improves the chance of recovery and - more importantly - ensures that the surgery ultimately is a success.
OUT OF THE OPERATING THEATRE AND INTO THE CAFE
These doctor-patient consultations are the nature of Data Cafés, the subject of this article. The premise is to have a consistent, fixed time slot to meet with users for an open discussion about the solution for training & feedback.
To give a concrete example for a Power BI reporting solution:
A weekly 1.5h Teams Call
Organized for a Sales Operations Power BI App
Run by the business sponsor of the app & team who built the reports & datasets
App end-users attend the call to ask questions and give feedback
The business sponsor announces new features & functionalities or coming changes
Demonstrations help users understand how to best use the App & Power BI reports therein
The Data Café would be run by the solution business sponsor together with the technical owner, whereupon they answer questions or share demonstrations & plans for the solution. It serves both a training as well as a change management function, but the most important point is to build a bridge between makers and users; to drive common understanding and eventually promote adoption.
WHAT VALUE DOES A DATA CAFE BRING?
A data cafe will improve not only solution adoption, but also the data skills and literacy of all involved. It will allow people to collaboratively create value from our organization’s data, and sow the seeds for long-term collaborations between data consumers & data creators. In effect, a successful data café drives & accelerates us toward a “data culture” in our organization.
WHAT DOES A SUCCESSFUL DATA CAFE LOOK LIKE?
On paper, running a data café is simple. All you need is an invite and a presenter. The reality, however, is that doing this successfully requires a lot of nuanced considerations. The reason is because a data café is only successful if it occurs frequently enough and with sufficient attendance. The Data Café shouldn’t be seen as a tool, but rather the central nucleus of a solution’s community of practice; it sets the standard for the data culture around the solution. A successful data café has the below characteristics:
Clear, enthusiastic communication
The overall purpose of a data café is to facilitate effective communication between makers & users. To this end, the person running the café needs to have good communication skills. They need to engage the people attending, and explain things so everyone understands. They also have to help the users in articulating their questions or needs, translating them for the group. While everyone in attendance has to try to communicate effectively, it’s the responsibility of the organizers to help make this go as effectively as possible.
Below are 3 examples of features that demonstrate this characteristic:Concise communication; users are not overwhelmed.
Engaging, charismatic presenters drive the energy & momentum of the data café
Concrete examples, visuals, demos or explanations to illustrate concepts discussed
Consistency in all things
When it comes to adoption, consistency is key. Running a weekly or bi-weekly data café and ensuring consistency in the quality, format, and organization will help build trust among the user community. When people know what to expect, they are also more likely to feel comfortable attending and sharing their own thoughts. It also makes it easier for them to understand the meeting and explain it to others so it propagates. Being consistent is challenging, but it’s all about building habits and putting in the sustained effort; it isn’t something that happens overnight, but rather stretched out over 3-6 months or more, at least.
Below are 3 examples of features that demonstrate this characteristic:The data café is held at the same day and time; maybe rotating alternately
The format, presenters, language and tone are always the same unless iteratively improved
It’s a “persistent campaign” where each café feels like a continuation of the last
Collaborative & open atmosphere
If no one is participating in the Data Café, or worse, if there is an antagonistic or even combative atmosphere between groups, it will not be valuable. It’s essential that the data café serves as an olive branch; a bridge between makers & creators. Doing this means ensuring that people can freely share their ideas and feedback without scrutiny or others reacting in a defensive way. People can ask questions without fear of looking ignorant or stupid.
Below are 4 examples of features that demonstrate this characteristic:Attendance / participation is consistent and high
Feedback and changes from previous cafés are regularly incorporated/demo’d in the solution
People can openly disagree without it getting uncomfortable or offended; it’s possible for people to disagree but still see the other person or group’s point of view.
Conflicts are approached, resolved or set aside in an impersonal and empathetic way
A clear value for attendees
This is the bottom line; it needs to be valuable for our company and our organization. It’s not sufficient to just say ‘we are improving our data literacy’, even though that’s true. There needs to be a hyper-concrete, tangible value each café for everyone in attendance. They should be able to walk away knowing something that will help them perform in or execute their tasks, or have contributed such that this will be the case in the future.
Below are 3 examples of features that demonstrate this characteristic:Discussions are about business questions & problems, not technical features / solutions
People regularly see their suggestions & feedback implemented in the solution; there is a communal sense of ownership over the data product, and not a “data hero” mentality of the technical solution owner.
Attendees find it a valuable use of their time; managers encourage direct reports to attend.
ADDITIONAL TIPS FOR A SUCCESSFUL DATA CAFE
-
Start talking about the Data Café with different users, and mention it in the landing page of your reports & dashboards. Create some buzz about it; maybe ask an executive working in the area of the solution scope to briefly mention it at a next meeting to get some top-down sponsorship.
This will ensure that it’s on people’s radar and they are not caught off-guard by the invite.
-
Send out a styled outlook invite that looks professional and attractive; something that grabs peoples attention. Put effort into the aesthetics and clarity of the communication. Make sure that someone can read and understand it in < 30 seconds. For example, you can use a lot of iconography, colour or separation of text in space.
-
Don’t force people to attend the Data Café. Ensure that it is an optional, drop-in / drop-out kind of meeting. If its held in person, make sure you have chosen a suitable location and room to accommodate everyone comfortably, but also so that people coming & going doesn’t become disruptive.
Make clear in the invite that it’s not mandatory and also not a formal training. Users should come prepared to engage, ask questions, and contribute with feedback… not just passively consume demos and training material.
-
Hold it over lunch with food, or after work with drinks. If that isn’t possible, then try to make it feel like a fun activity where people want to attend, and emphasize what they find important or interesting in a helpful & empathetic way.
If you’re looking for a more operational way to incentivize attendance, consider discussing options with the relevant HR department. It might be that people can include it in a kind of personal development plan or for their personal performance targets, for example, as a part of organizational training & upskilling initiatives.
-
Play agreeable music lightly while people are joining the call or the lobby, or make small talk asking people how they are doing, or discussing light topics. If you are the Business Sponsor or Technical Owner, you are responsible for the environment being a safe and healthy place where people want to be.
-
Try to ensure that everyone has a chance to speak. Don’t react - verbally or non-verbally - poorly to someone else’s opinion or question, and don’t tolerate toxic behavior in any shape or form.
Let people challenge existing dogmas or even best practices. Ask them for justification and reasoning, of course, but keep all discussions friendly.
Use “we” as the predominant pronoun instead of “you” or anything in third person.
-
Try to discourage people from working during the meeting or conducting other activities that take their attention, like using phones. This is mainly important if it’s conducted in-person, and is the value of the meeting being drop-in / drop-out; anyone should feel fine with coming and going as they need / want to.
When the meeting is done online this is less important unless people with their cameras on are clearly not paying attention, in which case it becomes disruptive.
To avoid situations where people are heavily distracted, it’s best to avoid creating strict rules. Instead try to understand why people are not engaged and think of ways to engage them.
-
If you are presenting, it’s important that you look and sound enthusiastic & charismatic. Even if this doesn’t come naturally, practice makes perfect. The reason why this energy is so important is because it is the lifeblood of the meeting, of the community. If you are consistently providing an energetic, enthusiastic environment that is open & welcoming to other people’s thoughts and discussions, people will be encouraged to participate.
However, if you are unenthusiastic or even sounding bored/cynical, it will drain the life out of the room and kill the atmosphere.
-
This is arguably one of the most important points. It’s absolutely essential that you do not become defensive when someone is criticizing a report, visual or dataset that you made. Even if they are criticizing the tool itself, try to ask them why they think that way and empathize with their frustration. Acknowledge their feelings and see it from their point-of-view; give them credibility. One of the easiest ways to build credibility with a user community is to demonstrate that your identity is separate from the solution; that you don’t take criticism personally and also aren’t attached to anything you made for them. This demonstrates that you’re dedicated to helping them and addressing the problem, and they will be more open to collaboration, honesty and solution-oriented thinking.
Then, once they have shared their criticisms, drive the discussion toward solutions. Try to identify what the problem is — independent of the person — and think together with the group about solutions. If the problem is too complex or nuanced, or it’s not being discussed in a helpful or productive way, then simply accept & thank the feedback and take it up after the meeting.
It’s also perfectly fine to acknowledge feedback and ask more time to digest it, afterward.
-
Inevitably, someone will question the numbers in the report or data solution, or bring up a topic like export to excel, particularly if it is disabled.
These are very sensitive topics and it’s important that you try to address them constructively:
Hear out the person. Let them explain the issue and where it comes from.
Acknowledge their feelings and their point-of-view. If the numbers are blatantly wrong, admit it; agree. If it’s not clear, make sure to reassure them that it will be investigated.
Ask them what they are doing with the data;
i.e. why do they need it in Excel, what do they do with it in Excel
or
i.e. Where they have obtained these other baseline numbersRequest to discuss it in a separate 1-1 call in order to give them the attention they deserve.
Do not waste time investigating data mismatches in the meeting, or arguing about sensitive topics like Export. Instead retain the integrity of the meeting while shelving the discussion for a more appropriate space. This makes clear that the issue is being taken seriously and you want to solve the problem, but also sends the message that the Data Café is not the appropriate space to discuss those things. The reason why not is because it can quickly unravel the trust and atmosphere, and when handling the topic on-the-spot, you won’t be able to properly investigate.
-
Like #9, this is also important for building credibility. If you are honest and admit when you are wrong or don’t know an answer, people will be more likely to trust you. They won’t feel like you are hand-waving their concerns or feedback, and will have more confidence in the answers you give.
To build trust, it’s important to accept that we are all flawed and make mistakes. This will also foster a culture where mistakes can be acknowledged, because they happen from everybody. Maybe you made a mistake in the report, but maybe a user made a mistake in their filter selections. That’s perfectly fine; it happens. It should feel that way during the Data Café.
-
When something is not possible or will be difficult to do, don’t beat around the bush and over-promise. Be clear about the effort to investigate or implement something, and be honest when the answer is “it’s not feasible at this time”.
However, try to stay solution-oriented; propose alternatives and try to solve the problem with other means.
-
This can be as simple as having a new filter available, or a formatting improvement. Having a few small announcements and updates every week will help reinforce that progress is being made and the solution is being cared for. This should be especially true if there is a large volume of feedback.
-
Even if you are having a very vocal, active Data Café, it doesn’t necessarily mean all user-groups are represented. Regularly compare the usage metrics with the Data Café attendance, and try to find active users who don’t attend the meetings to understand why. Ensure that everyone has a chance to participate and share their feedback, and try to find any ‘hidden’ groups that are not represented for any reason.
-
This is a more nuanced tip. It’s important to prepare the Data Cafés so they feel professional and well-organized; however, over-preparing makes them feel rehearsed, market-y, or rigid. Worse, if you over-prepare, you might have a mental picture of how the meeting will go that doesn’t fit reality. For example, you might prepare 30 minutes of demonstrations, but the end-users have a lot of questions. You want to get to the demos you prepare, so you cut the questions off with 30 minutes left and finish with the demos. This is a bad choice, because that cut-off feedback could have been extremely valuable. It’s important to be comfortable with ‘going off the rails’ a bit during these meetings; don’t be afraid to improvise.
This is implicit to having an open and collaborative atmosphere. It takes time to get there, but is very valuable, once attained.
TO CONCLUDE
We often like to think of our data solutions as tools that we pick up, use, and put down. In reality, the decision-making ecosystem is more like a living body. Removing an existing solution and deploying a new one is, in that regard, more like a surgical transplant. To ensure surgical success, we need more than skilled practitioners; we need adoption. We need to follow-up with the end-user community with rehabilitation, consultation & care. This will help our solution integrate into that ecosystem and enable our users to not just heal, but to become more efficient, more functional and more healthy as a direct result of our intervention.
