How your Current Data Strategy could be impeding growth
As a Business Intelligence Analyst, I often wondered, “How does a company’s Data Strategy drive growth?” Typically, vendors would provide the same kinds of answers:
- Better data visualization leads to faster insights and ultimately better decisions
- Big Data is the future, because more data is better
- Tap into Dark Data to uncover hidden insights
- Use Realtime Streaming Data for better in-moment decisions
Yawn. We’ve all heard these pitches before and working in BI I knew there’s more to better decisions than just more data, more dashboards and more visualizations, especially since everyone just looks at pivot tables. Not to mention getting to insights faster doesn’t ensure you’ll get to the correctinsight. Faster does not equal better.
The other thing to notice is that a lot of contemporary wisdom is just “industry jargon du jour”, which changes almost constantly. I wanted to go “beneath the veneer” to uncover the fundamental relationship between a company’s maturity with respect to its data strategy, and the effect of data maturity on growth.
What I found shocked me:
How could this be?
To understand how this is possible, we first need a definition of “Data Maturity.”
Measuring Maturity with the Dell Data Maturity Model
I’m sure every business student has seen this picture (or one very similar to it):
What I wanted to see was another curve that quantified the maturity over time of a company’s Data Strategy, superimposed upon this growth curve.
What I needed first was a way to measure Data Maturity, and then plot that over time. To do this I drew from the teachings of a leading tech company: Dell.
CIO.com has an article in which they define the “Dell Data Maturity Model” (or DDMM). Given Dell’s cachet this seemed like a good place to start.
For anyone who doesn’t eat, breathe or sleep the industry jargon these descriptions may be a little abstract. So, let me build on this by illustrating the pains a company feels just before it transitions from one stage of the DDMM to the next. If you find yourself nodding along with one set of pains more than any other you have a good idea of where you are in the Dell Data Maturity Model.
You may have noticed the gap between Data Proficient and Data Savvy is much wider as compared to the other stages. This wasn’t a formatting mistake. There are so many challenges involved in making that particular leap that, naturally, the gap is wider. Most companies never make the transition. In fact, a lot of businesses erroneously begin investing in becoming Data Driven (AI, Machine Learning, Data Mining, etc.) by hiring Data Scientists before achieving the Data Savvy stage. As a result, most Data Scientists are frustrated that they spend so much time cleaning and plumbing data because the foundation was never laid.
In Dell’s Model, Data Maturity grows in steps or stages. But the one dimension missing from this diagram is time. That is, how long does a company typically spend in each stage of the DDMM? This is an important piece that we’ll dive into right after we talk about the different stages of Growth.
Stage 1: The Startup Phase
The company has only a few people, anywhere from 1 to 3 employees and a few systems.
At this phase you’re building a car and trying to drive it at the same time.
Growth will be slow and data-driven decision-making usually takes the form of ad-hoc spreadsheets.
Stage 2: The Scale-up Phase
This stage is all about trying to scale-out your production and sales efforts and just get product out the door. People are buying as fast as you’re producing, and you have little concern for inefficiencies as you skip over bumps in the road.
Think of it this way: Your company is a brand-new sports car and you’re driving around with the parking-brake up and a parachute strapped to the back. But it doesn’t matter. Your growth engine is so powerful you’re just blazing past everyone else. But, a big enough pot-hole or bump in the road, and all that friction catches up with you.
Stage 3: The Enterprise Phase
The fun, youthful years of exponential growth have come a close. It’s inevitable but all high performing companies eventually settle into Sustainable Growth.
This isn’t a bad thing because your brand recognition also makes you King of the Hill. It just means growth over time will be linear, instead of exponential, and inefficiencies play a bigger role in slowing down growth.
Why Data Maturity Lags Growth over Time
This is where the rubber meets the road. How long does a company spend in each of these Data Maturity stages, and when does a “transition” occur relative to growth?
Let’s quickly define a “transition” as a step-wise improvement in a company’s Data Maturity. These are typically tied to events like hiring a BI Consultant, eventually a BI Team, or building a Data Warehouse. The purchase of technology is typically a part of all these transitions.
To track Data Maturity against Growth over time, we combine a few things:
- Company Growth in terms of Revenue over time (Figure 1)
- The Dell Data Maturity Model (Figure 2)
- Typical pains experienced by companies at each stage of their Data Maturity (Figure 3)
In my experience, these things don’t just neatly line up when you overlay them, and that’s the reason we find an interesting pattern in most companies.
If you add time to the Dell Data Maturity Model you end up with a Data Maturity Curve which roughly illustrates how long most successful companies stay in each stage of the DDMM. After overlaying revenue growth, in my experience, you see this pattern emerge:
Now, as you can see, Data Maturity grows in steps, typically the result of transitions caused by “precipitating events” (more on that below). The Data Maturity Curve isn’t so much a “curve” as it is a series of steps. If, for the purpose of illustration, we fit a curve to those steps and overlaid that, we’d see the following:
While I understand that (cross)correlation does not imply causation, what I expected to see was that as companies implement data assets, technologies or strategies, its Data Maturity Curve would lead ahead of its Growth Curve. And while higher Data Maturity doesn’t necessarily cause higher Growth, a leading Data Maturity Curve would at least have been in-line with my expectations. What I’ve found is the opposite.
Time and time again I have found this to be true of companies. A business’s Data Strategy (the different stages of the DDMM) is always behind the current state of growth. So, if you’re an executive leading a growing company, and reading this article right now, your company has already outgrown its current data strategy.
Why is this?
Because data-related pains don’t suddenly appear overnight. Unlike any other pain that physically prevents you from doing business (POS failures, internet outages, etc.), Data Pains are subtle, pernicious, chronic pains that compound.
It’s a lot like lower back pain, something many of us just learn to live with. It’s there. It sucks. But we get on with our day. Startups in the Data Aware stage barely feel it. But if left untreated that pain can show up in the worst and most unexpected ways, and you will pay good money to make it go away. Anyone who’s spent an evening lying on the couch due to back pain will understand this. It’s these sorts of “precipitating events” that motivate step-wise improvements in a company’s Data Maturity.
The above graphs are meant to illustrate a specific point: The maturity of a company’s data strategy lags company growth. So, you’re almost always behind the curve when it comes to your Data Strategy. But the keen reader should realize that, although we have graphed pain (and maturity) over time, they are not causally linked to time. They’re linked to growth.
This is the key distinction here. The pains themselves are tied to growth, not time, technology or the data itself. Untreated pains compound not over timebut over growth. Likewise, new pains occur due to growth, not the passage of time.
You may even call them Data Growing Pains.
Data-related pains tend to become unbearable at certain levels of growth. In the above graph we use revenue as a measurement of growth, but this is also very true of headcount, number of locations, countries in which the company does business, number of SKUs, and so on. This chart shows approximately where data pains become almost unbearable for Apparel businesses:
- At around $10M revenue and roughly 10 people at head-office, integrating data from multiple systems becomes unbearable. The company needs classic Business Intelligence that can consolidate data from multiple source systems.
- At around $150M-$200M revenue and close to 100–200 people, new pains around people and the process of using data to make decisions begin to surface. The company needs a Decision Support System (enabling Centres of Excellence).
- Coming up on $1B in revenue and crossing 500-people, companies look towards Big Data, Machine Learning and AI (Data Science) to create micro-optimization at scale: real-time recommendation engines that cross-sell products online, or chatbots that improve customer experience and open new channels.
Different industries will have different milestones in terms of revenue or headcount at which data pains become unbearable, but the general shape of the Pain vs. Growth curve will be the same. The pain will drop soon after a company undergoes a “transition” to move from one stage of maturity to the next.
What happens if you never invest in your Data Strategy? Some executives believe the first step of data integration “is a showstopper,” leaving them stuck at a nascent level of Data Maturity despite their growth. This is a bad situation to be in, because not only will you inherit all the new problems associated with growth, the original problems which went unsolved just get worse.
Fortunately, most executives realize the need for Business Intelligence. Unfortunately, due to a lack of available technology, industry hype around Big Data and AI, or poor guidance from industry experts, most companies never cross the chasm between Data Proficient (having a BI system) and Data Savvy(Decision Support System enabled Centre of Excellence). In fact, some companies try to skip the Data Savvy stage entirely and go straight to Data Driven (using Data Science).
As a result, most companies’ Pain v. Growth curve actually looks like this:
The silver lining here is that, while the pain will continue to worsen with growth, it doesn’t get exponentially worse. Most companies think that the solution here is to invest in AI and Big Data, and while they certainly have the data quantity to do this, they don’t have the foundations in place to act nimblyon the insights (or opportunities) that come from Big Data or AI.
It’s also important to realize that if a company never expands beyond a certain level of growth, it’ll never feel the (new) pains that accompany that (new) growth. If the business never adds countries, stores, SKUs, warehouses, customer types and so on, they’ll never need to transition to a higher stage of Data Maturity. This is important for any executive who believes he or she can simply “deal with it later” and “focus on other priorities.”
In a future post we’ll describe what these pains look like and how they manifest in the day-to-day operations of a business.
How Process can be a Competitive Advantage
You may be thinking, “OK, Company Growth causes Data Pains which, when solved, result in step-wise maturity in our Data Strategy. But who cares? We’re still growing in spite of that.”
Maybe you’ll continue to grow. Maybe not. The question you should be asking is, “How much money are we leaving on the table by not solving these problems?” Another question to ask is, “How long before these inefficiencies start to catch up with us?” and “Will it cause a premature slow-down in growth?”
The best question to ask is,
Dashboarding and Business Intelligence have become commoditized and easily accessible for most companies. This is a good thing, and especially beneficial to companies on the smaller side. Tech giants are filing patents on hardware and software that make it faster and more economical to crush Big Data and incorporate AI. This is useful to ultra-big enterprises.
But what about everyone in-between? If you review the pains listed in this article, you’ll notice a lot of them relate to the process of using data to make decisions. Collection, cleaning and integration play a (very big) part in the early stages of growth, but process, people and business unit integration play an equally big part in later stages. While most of the data-industry is focusing on the sexy things like AI and Big Data, the truth is most companies will have problems with the process of making decisions with the data they currentlyhave long before they are ready for AI and Big Data. Many large enterprises never truly make the leap between Data Proficient to Data Savvy, with leadership choosing instead to hire Data Scientists and invest in sexier technologies.
This presents a huge opportunity for Scaleups.
When you think about patents you think about technology as a competitive advantage. Not many people think about process, which is as equally valid a thing to patent. And while I’m not saying you need to run out and file a patent on how you make decisions with data, what I am saying is that if you can nail a process that allows you to be nimble with your data early in your growth curve then you have a huge advantage over larger companies that didn’t.
A Scaleup’s major competitive advantage is (still) having the ability to jump into a meeting room, pull up some numbers and make decisions. If a company can retain that agility throughout its growth it can continue to run circles around its larger rivals. The net effect is reducing the lag between the Growth Curve and the Data Maturity Curve and, ultimately, the data pains associated with growth. In future we’ll dive deeper into how this can be done.
- The maturity of a company’s Data Strategy (its Data Maturity Curve) often lags its Growth Curve
- Better visualizations could mean faster insights, but faster does not equal better since you can get to a wrong insight just as fast as the right one.
- Data pains are directly tied to growth; not time, technology, or data itself.
- To cross the chasm between Data Proficient (Business Intelligence) and Data Savvy (Decision Support Systems), focus on curing inefficiencies in people, process and business unit integration.
- Don’t try to skip the Data Savvy stage by focusing too early on Big Data and AI (even if you have the data quantity and resources to do so). To become Data Driven, lay a solid foundation around communicating and actioning on data so that you can get the most out of your future Data Scientists.
- Most of the data-industry is focusing on the sexy things like AI and Big Data. The truth is companies will have problems with the process of making decisions using the data they currently have long before they are ready for AI and Big Data.
- Focus on best-practices, and the technologies that enable them, to improve efficiency in the decision-making process before looking at Big Data, AI, or any tech hype.