Data science is becoming a reality for change management, and although it may not have arrived yet, it is time for organizations to get ready. The companies best positioned to change in the next decade will be the ones that set themselves up well now, by collecting the right kind of data and investing in their analytics capacity.
The key to building predictive models is knowing what you want to predict and collecting large and diverse data sets that may enable you to do so. Although predictive models for change management are still a ways off, organizations can get themselves on the right path by adopting the right tools and capturing the right data. We see five no-regrets steps that organizations can take:
Start Using Digital Engagement Tools
There is a new generation of real-time employee opinion tools that are starting to replace old-fashioned employee opinion surveys — tools that tell you far more than just what employees think every year. These tools have obvious relevance to change management and can help answer questions like: Is a change being equally well received across locations? Are some managers better than others at delivering messages to employees?
We are working with a large travel and tourism firm to introduce a system for real-time employee feedback. This is giving us the opportunity to experiment with different change strategies within chosen populations in the company. The real-time feedback means we will learn very rapidly how communications or engagement tactics have been received, thus optimizing our actions in days rather than weeks, as might be the case with traditional approaches. This data can then feed into a predictive model, helping us know with precision the actions that are going to accelerate adoption of a new practice, process, or behavior by a given employee group. Commercially available tools, such as culture IQ polls, sample groups of employees through a smartphone app on a daily or weekly basis to generate real-time insights. Waggl.com goes further, creating an ongoing conversation with employees about a change effort, allowing change managers to tie this dialogue to the progress of initiatives they are undertaking. These tools can already have a big impact on change programs, but the data stream they create could be even more important as we learn to build predictive models of change. Deploying them now is critical to ensure success with data-driven change initiatives in the future.
Apply Social Media Analytics to Identify Stakeholder Sentiment
Change managers can also look beyond the confines of the enterprise for insight about the impact of change programs. Customers, channel partners, suppliers, and investors, to name just a few, are all key stakeholders for change programs. They are also more likely than employees to comment on social media about changes a company is making, thus giving potentially vital insight into how they are responding. At EY, where some of us work, we have developed a tool for social media analytics called SMAART that is able to understand sentiment within consumer and influencer groups. In a project for a pharmaceutical company, we were able to isolate the specific information sources that drove positive and negative sentiment toward the client’s brand. We are now starting to apply these techniques to understand the external impact of change efforts, and it’s a simple leap to extend these techniques within the enterprise. Advances in linguistic analysis of text mean we can now capture clues about behavior from people’s word choice; even the use of articles and pronouns can help reveal how someone feels. Applying these tools to anonymized company email or the dialogue on tools like waggl.com will give fresh insight into change readiness and the reactions of employees to different initiatives. And the insights from analyzing internal communication will be stronger when combined with external social media data.
Capture Reference Data About Current Change Projects
Organizations often seem obsessed by measuring fractional shifts in operational performance, capturing data on sales, inventory turns, and manufacturing efficiency. However, when it comes to change, few track performance from project to project beyond knowing which ones met their goals. Although projects have unique features, there are many similarities between process improvement, system change, M&A, and reorganization projects. There are opportunities to capture information about the team involved, the population engaged in the change, how long it took to implement, what tactics were used, and so on. Building a reference data set like this may not yield immediate benefit, but as the overall data set grows, it will make it easier to build accurate predictive models of organizational change.
Use Data to Select People for Change Roles
For decades companies have been using data-driven methods to select candidates for senior roles. And today some businesses such as retailers are using predictive analytics for hiring frontline staff. Applying these tools when building a team could both improve project performance and help to build another new data set. If every change leader and team member underwent psychometric testing and evaluation before the project, this data would become variables to include as you search for a causal model on what leads to successful change projects. This can even be extended to more informal roles like “change agents,” allowing organizations to optimize selection based on what they know about successful personalities for these roles. Along these lines, California startup LEDR Technologies is pioneering techniques to predict team performance. It integrates data sources and uses them to help teams anticipate the challenges they may have with team dynamics, so they can head them off before they occur.
Build a Dashboard
We imagine each organization having a bespoke dashboard, developed in partnership with the firm’s leadership team, reflecting their priorities, competitive position, and future plans. In this way, dashboards can bring insight to specific transformation investments the organization is making. Much of the data that makes up these indicators is already available today but is not being collected. One client of Change Logic’s has built a dashboard for identifying recruitment and attrition in must-win talent populations. It’s not as sophisticated as some of the models we expect to see, but nevertheless it is teaching the executive team to use data to inform people-related decisions.
It will take time to build these sorts of tools. We believe organizations should start building dashboards now and, where possible, automate them. Today, change dashboards are vulnerable to version control issues, human error, and internal politics. Automating the dashboard can make it more transparent and objective.
As organizations collect more data and build more-accurate models, change managers will be able to confidently use them to prescribe strategies to enable organizations to meet their goals. Which stakeholders are involved? What approaches work with groups that share these characteristics? What risks are associated with programs that share these features? What are the techniques that accelerate delivery of business benefit, and what are their relative costs? What is the cause and effect of specific types of investment — for example, leadership development, large group events, and communications cascades? These are all questions that will be answered with data and that will inform customized transformation plans.
Developing these sorts of metrics will not be quick or easy. These are not one-and-done installations, but rather multiyear commitments to capture data, build models, and refine dashboards. Establishing reliable data sets with which to work takes time. Data quality is an issue everywhere; so too is the need for a common data language that allows organizations to know they are measuring what they intend to measure. This has been a problem for data analytics in other fields; there is no reason why change management will be any different.
Although it will take time, we will finally be able to close the causal loop and make reliable predictions for how an action or initiative in a change program will shift a given metric. This will move investment in change from an act of faith to one of data-informed judgment. Change management will move from a project-based discipline struggling to justify adequate investment to one that is advising on business outcomes and how to deliver them. This should lead, at last, to a decline in the one metric about change programs that we all know — failure rate. And along the way we may finally solve the great puzzle of why so many transformation efforts fail.
Post a Comment