Analytics as an Agent of Change

Analysts are trained by most schools to focus on databases, statistics, machine learning models, programming and visualization. These are all incredibly important but the hardest part about analytics is that it's often not about any of these things. To really leverage data, means change. And no matter the size of your company, its industry or how mature they are, change remains universally hard.

Analytics & Change

Regardless of the work you're doing in analytics, success means change.

• If you're bringing an insight about some new consumer behaviour to a stakeholder, you want that to bring a change. You want them to change how they market, sell or even what products they focus on.
• If you're building a new model to automate a part of customer service, you are actually changing the jobs of your customer service team.
• If you're building a BI tool, you want to change the daily life of your user so they make this part of their lives.

The more meaningful your analytics' project, the greater the change you are making. If you complete a project and can't point to change, was it really successful?

The more meaningful your analytics' project, the greater the change you are making.

Change Is Hard

But, change is hard because humans are ... well human. Universally, we have emotions, fears, biases, egos and beliefs. Good data and sound logic may not be enough.

• Maybe someone feels threatened that your telling them what the customers need?
• Customer service reps may see your AI as a threat for their job.
• Maybe your BI customer, simply doesn't see the need for the data or the tool does fit with how they want to see the data.

We want to believe that every data insight and every ML model will be cherished and valued. That stakeholders will jump at the opportunity to leverage them. That may be true some of the time, but it won't happen 100% of the time.

If you deliver the best data product or best insight in the entire world and no one uses it, it's valueless. It may have the potential to be valuable, but until it's used it's valueless. As an analytics professional, it can be heartbreaking but it's the truth.

If you deliver the best data product or best insight in the entire world and no one uses it, it's valueless

Sidebar: Analytics as a Data Drivethrus

There's an interesting dynamic I've seen play out over the two decades I've been leading analytics. There's a balance between being responsive to stakeholders and being a change agent.

Being slavishly responsive to what your stakeholders ask is - in my opinion dangerous. It can result in the team becoming nothing more than a Data Drivethru. A stakeholders sends a request for data, it's dutifully pulled from the database and then provided. It's easy to say that the risk of failure from change is very low, but so is the value. I'd argue that analytics must drive change and avoid becoming the McDonalds of Data.

The opposite can also be true. Trying to drive change all the time and being completely unresponsive can made you vulnerable - frankly push to hard without support and it may cost you your job. You need the support of stakeholders and their trust especially when you are trying to drive change.

So the best position is to walk a balance between the two. Build up credibility by being responsive and use it to drive change.

Driving Change via Analytics

So, let's talk about how to help ensure that you can take your amazing work and have a better chance of getting it

First, let's assume some basic prerequisites. These aren't really the heart of change management for analytics but they sure can block you.

1. You need to be accurate and reliable. Low credibility will make your life much harder.
2. Your presentations and communications are clear and easy to understand.
3. Your projects can be connected to the metrics and issues your company cares about.

So, this leads us finally to change management. If your team's success or failure is measured based on how you're impacting the organization, it's your responsibility to manage the change within your organization.

If your team's success or failure is measured based on how you're impacting the organization, it's your responsibility to manage the change within your organization

So how to do you manage change? There's literally books on the topic and you should probably read a few. But I'll give a few suggestions.

First, you need to ensure you have the support of your most senior stakeholder. Projects fail if they don't have the support of senior leadership. So, here's a great chance to cash in those chips you've earned by dutifully being responsive.

Second, you should apply a formal approach for large risky projects. ADKAR is a very simple framework that you can leverage.

Awareness: Create awareness of the benefits of changing and the need for change in order to move the organization forward.

Knowledge: Provide knowledge and training on the change process and how it works.

Ability: Provide the resources and support to enable employees to be successful in the change process.

Reinforcement: Enhance internal and external communication and engagement to reinforce the desired behavior.

Acceptance: Monitor progress and offer incentives to increase acceptance and support for the changes.

Every stakeholder affected by change needs to go through these steps. Some stakeholders may go through that very quickly without any support. Other stakeholders may need help and support to get through them. You spent months cleaning data and writing code, why not spend some time to ensure your project gets completed.

Lastly, measure the projects impact. Be as scientific and honest about how your changes impacted the organization as you would anything else. Then share the results transparently. This creates immense trust that your team (and projects) are creating value and that someone is watching it after it launches.

Analytics requires a lot of skills and it sucks to add another skill to the list but it can be difference between success or failure. To many good ideas and projects die at the final mile.