Why Analytics Should Be Standardized
...and how to do it
Hello and welcome to the latest issue of Data Patterns. If this is your first one, you’ll find all previous issues in the archive.
In the last few issues we’ve started talking about making sense of the data landscape using Wardley Maps and figuring out why there are so many problems in the field of data.
In the last issue I claimed that there are only two core problems as the root cause of all the other undesirable effects in the field of data:
The value of analytics is opaque/unclear
Not everyone knows how to use analytics to improve the org
In this issue I will offer one potential solution to both. I decided to change up my writing style to see if I could convey the same message using the Socratic approach. Therefore what follows is a (not real) conversation between an executive and a consultant who is explaining the idea.
Executive (E): “Ok, you’re saying there’s only two core problems in analytics. It seems to make sense. So what do you propose the solution would be?”
Consultant (C): “Yes, if executives understood the real value of analytics and knew how to use it to improve their business, that would lead them to prioritize data quality and metrics. If these metrics were standardized, it would make everything easier.”
E: “What do you mean by standardized metrics? You mean like the same metrics everywhere?”
C: “In a way yes. A good chunk of metrics can be standardized for all businesses based on industry best practices. That enables you to get value out of analytics quickly with minimal resources.”
E: "But how's that possible? Aren’t businesses different? Our business strategies and many of our processes are unique. We're not just another company in the SaaS space."
C: "True, every business has unique elements, but there are many operational aspects that are strikingly similar across businesses. For example, you're in the B2B SaaS business, right?"
E: "Right."
C: "What’s your business model? How do you generate revenue?"
E: "We use both a recurring revenue subscription model and a usage-based model."
C: "Great. Now your main competitors, how do they operate financially? What’s their business model?"
E: "Oh I see what you mean. They also have a subscription and usage based model. But our features are different, our pricing is different. Doesn’t that make us unique?”
C: "Those parts are always unique. When it comes to analytics, though you can standardize a significant chunk. Around 70%-90% of the metrics are typically common in all B2B SaaS businesses."
E: "70% to 90%?! That seems high. Aren't our specific marketing campaigns or unique sales processes going to significantly impact these metrics?"
C: "Let’s think about it. For example, how do you currently measure your marketing efforts?”
E: “Well we have customer acquisition cost (CAC) that we base on first touch attribution. We also have several funnel conversion rates we track across multiple channels”
C: “If you went to work for another B2B SaaS how do you think they would measure their marketing efforts?”
E: “Ah ok I see your point. They’ll also use CAC. But it still seems high though. What about business strategies? Aren’t those unique?”
C: “Yes, that's what makes you unique. This is more about industry best practices. For example when you built your sales team, did you start from scratch, trying random things to figure out what worked or did you hire industry veterans with a lot of experience?”
E: “We hired experienced sales executives and they indeed start with best practices. But these were heavily modified based on our unique needs and our specific client base."
C: “Right, and for marketing?”
E: “Same thing.”
C: “What were you hoping to accomplish by bringing a seasoned executive on board? What were you hoping to avoid?”
E: “Well, we were hoping to use their knowledge and expertise to build the best team possible. We wanted to do it quickly and avoid common pitfalls.”
C: “Doesn’t that mean that there exist certain best practices, of what works and what doesn’t for every aspect of the business like sales, marketing, product development, engineering, operations, finance, etc?”
E: “Sure, but isn’t that unique to each industry? For example, what works in marketing for a B2C business is completely different from B2B marketing.”
C: “Yes that’s true. The approach is probably different, the methodology could also be different but the fundamental function is the same.”
E: “What do you mean?”
C: “For marketing, for example, you’re still spending money on campaigns to generate demand and supply the sales team with more leads. And when you spend money to do something it’s only natural to know how effective your efforts are. Whether it’s B2B or B2C you’re still measuring the same thing: how much you spent and how many customers you got.”
E: “I see. But my customers are different, they come from different segments than B2C. How does this standardize?”
C: “The segments are different, the campaigns are different, the costs are definitely different. What’s the same is the fact that there is a cost, there is a campaign, there are segments. You could ask a marketing person in an B2B, B2C or eCommerce business about their segments, their campaigns or their CAC and they’ll know exactly what it means. Therefore CAC can be standardized. The same applies to many other metrics.”
E: “Ah ok, that makes a lot of sense. You’re saying that while my sales or marketing efforts are different, how they’re measured is the same across many businesses, right?”
C: “Exactly! At a more abstract level there are proven ways to measure the performance of every aspect of the business. And that’s where the problem lies.”
E: “What problem?”
C: “Well the way data teams operate these days is as if these standards don’t exist. They walk into a new company and have to build everything from scratch. From data modeling, to metric definitions, to dashboards and reports, everything is custom built and unnecessarily unique.”
E: “Wait, don’t they know what to build? If I hire an experienced data scientist I expect them to know what they’re doing.”
C: “They know what they’re doing from a technical perspective. They can write code that prepares data, calculates the metrics and builds dashboards. They understand data visualization and statistics which allows them to come up with insights and offer recommendations on what actions to take. But when it comes to the methodology of which metrics to implement or how data should be collected in order to make modeling easy, they’re reinventing the wheel every time they join a new company.”
E: “I’m not sure what you mean. Shouldn’t they already know what to do?”
C: “Because there’s no agreed-upon standard for how companies are measured, aside from GAAP accounting, every company implements their core metrics from scratch. Analysts have to talk to each stakeholder to get their definitions of the metrics, understand how raw data is structured and then write the necessary code. Often there are conflicting opinions on even basic things like what’s a sale or what’s a customer.”
E: “Hmmm, ok.”
C: “That’s a lot of wasted effort. The time to value of analytics increases, costs increase and projects are abandoned. You often get metrics duplication, like having multiple copies of revenue that don’t match. That leads to confusion, frustration and eventually loss of credibility in the data team.”
E: “Yes I’ve seen that happen before multiple times. So what are you saying; you have a solution?”
C: “Well yes. That’s what SOMA does. It offers both a standard definition of the most widely accepted metrics for each aspect of the business. It also offers a standard way to model data through activities. It works like a checklist. Your data team can walk into a meeting with a stakeholder and use SOMA as a way to check what’s already been implemented and what remains.”
E: “Oh that’s interesting. That way everyone agrees on key metrics like revenue. But what about our unique needs? I want to make sure they’re not overlooked.”
C: “That’s the idea. Once adopted, a good portion of your analytics can follow the industry best practices, which speeds up implementation and reduces cost. Analysts can then focus on the uniqueness of the business, which is often where innovation comes from."
E: "That's a compelling argument. If this approach allows us to move faster it's definitely worth looking into. How do we get started?"
Special Offer
If you work in a B2B SaaS business and are interested in implementing SOMA, hit reply and let me know. We’re looking for companies to partner with to help advance SOMA development. It’s an open source standard. You can find all the details here.
That’s it for now, I’ll write more about this topic because it’s of very high interest to me and I hope to you as well.
Until next time.