Analytics data strategy is not just about technology, it’s also about people and processes. Tools are getting easier and easier to use every day, yet adoption rates and quality of analytics output are often stagnant. Why is that?
Well, it is usually because companies over-invest in technology infrastructure, but underinvest in the other pillars.
Traditionally, it has always been about item 1, the underlying technology. But in truth, analytics modernization is really all about items 2-5, adoption, investment in people & processes, and being human centered.
Modern enterprise data strategy needs to have modern analytics.
In this guide, we’ll help you get started on your modern analytics journey toward modernization by uncovering a few common roadblocks and diving into five doable steps to get you to analytics modernization faster than your competitors.
Analytics modernization can go by other names, including analytics transformation or business intelligence transformation. In all cases, one of the core goals is to increase analytics adoption across all levels of the organization–thus making the organization truly data-driven.
Before we begin, it’s important to think about analytics modernization and analytics adoption as a journey and not a discrete benchmark of “being there.” There are many reasons for regression, some which include staff turnover and lack of trust with data sources. Maintaining a data-informed organization takes consistent effort, and it is possible for organizations to regress in their adoption. This is a key point for anyone just beginning their transformation.
For most companies, analytics adoption is typically part of a larger digital transformation roadmap. The instructions on the roadmap often contain little direction from senior leadership noting how to execute analytics modernization.
Don’t let this lack of direction frustrate you! This happens with the majority of organizations. This is an opportunity for you to set the agenda for how to make your organization successful.
One final point: remember that every organization has different starting points, endpoints, paths, and current states. And despite great marketing efforts by some organizations, almost every organization in the world is still trying to figure it out.
Before we dive into the five steps to analytics modernization, let’s call out three symptoms that commonly plague organizations in the early stages of analytics modernization.
Shelfware is the lack of use and adoption of the technology already available in an organization. In the early stages of analytics modernization, pressure from senior leadership may be focused on getting employees to use the technology already available to them. This is the result of investments into the technology without having known what the demand would be. For instance, there could be an under-utilization of data in a reporting database (semantic layer) or in the business intelligence and data visualization tools.
It’s easy for leaders to fall into this trap. Leaders may start with purchasing technology because it was fit to solve a single problem. But as other problems unfold, leaders assume the same tool can solve all those other challenges. But there’s a lack of fit to those solutions.
There is no doubt that shelfware is a very bad condition for an organization to face. We believe that shelfware is a symptom stemming from a combination of challenges, including data literacy of an organization, talent development, internal partnerships, use cases, and backlog management. In many cases, the challenges can go even further to the data itself.
There will always be different viewpoints on a topic within an organization, particularly on actions and next steps. Ideally, the direction taken, regardless of viewpoint, should be informed from consistently-structured data.
Organizations in the early stages of an analytics transformation will often find employees challenging each other on the fundamentals of the insights themselves. This is because data are not organized in a fashion that makes it easy for analysts to derive consistent conclusions.
That is to say: one analyst might come up with completely different numbers than the next. Many times the challenge is simply how complicated the data are to work with.
These challenges can be traced back to organizational data literacy, federated access to unified data sources, and harmonized data sources.
Underutilized reports doesn’t mean that employees aren’t accessing reports. In fact, there might be several high-use reports. Instead, underutilized reporting means there are many reports with little to no use.
Organizations that have not started their modernization journey often hear from their users that they can’t find the reports they want and that there are too many reports for them to make sense.
That’s not to say that each report didn’t make sense at the particular time that it was created. Often reports are created with the best intentions. But reports should be living, breathing products.
These challenges can be traced back to challenges in governance and enabling self-service analytics. By enabling self-service, users can get what they need when they need it, by creating it on their own. And when the report is no longer used, it is removed from the environment.
Now that we’ve given a brief (but thorough) overview of analytics modernization and what can slow it down, let’s dive head-first into what steps you can take to fast track your way to analytics modernization.
Feel free to jump ahead to any of the various steps:
One of the most commonly-asked questions of data leaders is: should we be using a top-down or bottom-up approach for scaling? There is only one way: top-down. You absolutely need to have your senior leadership buy into the process.
The best way to get buy-in is by first finding leaders who are interested in partnering with your teams, developing their use cases (we’ll talk about this more in the next step), and elevating their successes. The more partners you can elevate, the more successful you will be overall.
When looking for the right partner, don’t overthink it. Work with teams that already look to get the most out of their data. This might be a leader who has a strong understanding of how to use data, or it might be a team where they have a strong individual contributor that can go far. Ask yourself: which existing relationships can you both leverage and highlight?
Of course, you’ll need to broaden your relationship within the organization. This is where you’ll want to do a roadshow highlighting the excellence of other teams and helping promote the art-of-the-possible with those other teams.
If you are struggling with getting buy-in, then focus on building deeper relationships with successful teams. The further you can take them, the greater proof you will have of how far you can take the team on the roadmap.
Before you can execute these use cases, it’s best to work with your chosen teams to develop a strategy that prioritizes which centralized data sources will become available first. After you identify the use cases and build with what you have, don’t worry about having a fully functional data pipeline.
While you are working to centralize those data sources, you should use analytics processing automation tools like Alteryx, Tableau Prep, or Microsoft Power Automate to create high-fidelity prototypes of the data sources. These tools are also great for this process because they are self-documenting and can allow data engineers to quickly understand the business logic and translate it into a formalized table in a data mart.
Critical to establishing buy-in is creating steering committees of cross-functional senior leaders that are provided bi-weekly updates to the progress of any deliverables. The steering committees go beyond status reports: use them to help drive the direction of your modernization efforts.
The key amongst leaders is to create competition. These bi-weekly meetings highlight the successes. During these meetings, it is important to show how each unit is progressing in terms of key metrics, like tool adoption, time savings as a result of automation, and cost savings as a result of automation.
As mentioned above, one of the greatest challenges to gaining momentum for your business intelligence modernization is getting stakeholder buy-in. One of the underlying challenges of gaining buy-in is measuring the return on investment (ROI) of the projects completed.
When automating an analytics process, employees should be saving time, saving money, increasing customer insights, or unlocking new revenue streams. There are, of course, other ways of getting a ROI. Of these, the most common will be time savings, which should mean increased productivity.
To solve this challenge, it’s important that you sit down with each business leader (and their teams) to define each use case where analytics could be automated. As you identify each of these use cases, you should determine:
It’s important that when you identify the potential impact of a project, you explicitly state the value of the project and note the estimated time or cost savings. After you have defined the use cases for a business unit, you will need to build a formal product backlog by adding these into a centralized project management software. This backlog should be available to the business unit you are supporting and also available to your team.
After you define the use cases and build the backlog, you should immediately start creating the analytics solutions. By completing the top use cases, you will be able to quickly quantify value and gain support from the leaders of that business unit.
If you don’t have the resources to support it, then it’s important to find strong outside partners who specialize in this work. This will allow you to capture the value and advance the analytics transformation forward.
An early challenge you will encounter in the beginning stages of adoption is the current technical and data skills of employees at your organization. Just to be clear, every organization faces challenges in the data literacy of their employees–no one has this solved perfectly. But a prominent lack of data literacy will absolutely impede your modernization efforts.
You must separate data literacy from the technical tool upskilling that is necessary for your workforce.
Your entire organization needs to understand the basics of working with data: how it is structured, how to do simple analyses, and how many of the manual processes they are completing could be automated. They don’t need to know how to do it yet, they just need to be aware of how that could happen, and how valuable these skills can be to your organization as it potentially increases the capacity of each employee.
When an organization is data literate, it can make decisions grounded in the data. When decisions are grounded in data they are facts. But there’s more to it. When an organization is data-literate it can explore data in new ways to uncover new and more powerful insights.
Be sure to check out Datacoach, an analytics platform that helps organizations transform the analytic capabilities of their team using modern video lessons and one-on-one coaching from renowned experts.
With the tool or technical upskilling, this is the actual learning of the tools needed to drive analytics modernization of your organization by employees in the business.
For upskilling, find and curate content that can help employees learn how to develop with the tools. Don’t waste resources creating the content yourself. Also, ensure that the content covers relevant topics to your users. If your organization is learning an analytics tool like Tableau, don’t invest in training that just shows them how to make certain chart types. Invest in training that pairs the training with content that shows them how to perform real analytics challenges that are relevant to your organization
You’ll want to find an evangelist– or multiple evangelists– who can help coach employees and act as a product owner for the technology. This person needs the technical skills and the emotional intelligence to support individuals. Additionally, this person should be dedicated to the role. Note that as you mature further, there may be more than one person in this role, depending on the size of your organization.
If you feel like this person may be you– we highly recommend that you do not take on this role. You’ve got larger challenges across the organization to solve.
This evangelist is going to be the tentpole that drives the overall maturity of tool adoption–so the greater the skills this person has, the better. In many cases, this role is outsourced to consultancies who are specialized in the use of the technologies until the organization reaches scale.
The best action you can take to support learning is to combine the training resources with the use cases identified in the backlog. Have employees build their skills by developing analytics solutions directly from the backlog while learning the tools. After they complete the training and the backlog item, you can highlight the outcome in a steering committee of senior management.
It’s also essential that you consider data and technical skills on a continuum rather than discrete yes/no. Eventually, you’ll need to articulate what data and tool skills are necessary for key roles across the organization. Work with hiring teams to identify the true skills needed from new hires and what support you can provide to them during their onboarding process to help skill up during their first months on the job.
As individuals gain new skills, offer an internal credentialing process that highlights the growth of individual skills and teams. By gamifying these credentials, it will create a buzz amongst employees and help your team gain traction across the organization.
Building a strong learning culture is critical to the success of analytics modernization. It has been linked to increased morale and retention rates across organizations. But remember, you are increasing the skills and abilities of your workforce, so you should also have a plan on how to retain the most successful people who are highly engaged in the transformation process, as their new skills will be extremely marketable.
As mentioned before, one of your greatest obstacles will be competing narratives. This will come from having multiple data sources that offer the “truth.”
To solve this challenge, you’ll need to develop a data mart (now sometimes called a data lakehouse, if you are on the cloud) that defines business logic and makes the data easily accessible to business units.
Arguably one of the most overwhelming challenges will be the number of data sources that will need to be defined. Don’t worry about having all of them complete at once. Have your teams work in an agile-like manner and release data into a data mart.
You will undoubtedly be asked to scale this process. Depending on your funding structure, don’t be afraid to ask for more funds to either bring on full-time employees or a consulting partner.
Making data accessible to business units is critical for getting their buy-in. The faster you can earn the support, the easier it will be to scale and ask for continued support.
Remember, when solving this challenge, you will be forced to solve three questions:
As you begin to build your centralized sources of truth, your data will ultimately run into some ever-present issues: changes in business logic and the need to bring together or harmonize seemingly disparate data sources.
These requests will come early and often. Data sources that were defined and developed change. Rules don’t stay the same. So it’s important not to meet these requests with a sense of frustration. Just remember: they are ever-present in every organization. Instead, it’s better to acknowledge that the requests are a sign of the process working, as business units are submitting requests to you and your teams.
Develop a sustainable plan for managing data requests immediately.
This will allow you to clearly articulate to your stakeholders when you will expect to complete the task and how long it might take.
If there is a significant backlog in requests and you are lagging behind, it offers you two talk tracks for leaders:
The steps discussed are just the beginning of a much more nuanced conversation.
It is clear that effective analytics modernization is more than training users and building centralized data sources. It is a multi-faceted transformation project that requires a blend of technical and change management skills.
Building this capability within an organization requires a complex approach combining change management and product management skills with data and technology savviness.
The first step in effective data strategy is focusing on people and processes, not just technology. This, at its heart, is the purpose of analytics modernization.
If you’re looking to get a jumpstart on your data modernization efforts, or need a reliable consulting partner to help along the way, we’d love to talk! phData is passionate about helping businesses of all sizes succeed with their data and analytics journey.