Traditional vs. Minimalist. Which Analytics is right for you?
What's in this for you?
There are so many tools, technologies, and buzz words in data and analytics that it can make your head spin. The challenge is really understanding which of these things is right for your business. In the early days companies pick the tools before they pick the strategy. This usually happens because departments will buy their own tools independent of one another. This is a very “bottom-up” approach that can take a company down an expensive path.
So in this article we’ll talk about two, high level approaches that companies can take when embarking on their data strategy: The Traditional Approach vs. a Minimalist Approach. We’ll define what these are and help choose between the two.
Traditional vs. Minimalist Analytics
Traditional. As you might have guessed the Traditional approach to analytics has been around for decades. At a very high level, you do the following:
- Build an in-house, full-time Business Intelligence team from the ground-up
- Purchase and integrate off-the-shelf software that you customize
- Hire software developers to build custom data applications, or implement open source tools, to fill gaps in off-the-shelf tools.
Businesses that choose the Traditional Approach typically do so because they have a very strong need for something highly bespoke and customized.
Minimalist. The Minimalist Approach to analytics, unlike the Traditional Approach, is a fresh and disruptive take on building an analytics offering. At a high level, the Minimalist Approach requires the following:
- An outsourced, fractional Business Intelligence Team
- An all-in-one Reporting, Planning & Analytics (RP&A) platform that allows you to collect, clean and standardize your data.
- Favour configuration over customization to keep costs down and implementation times short.
This often leaves leaders asking – which one is right for us? It’s easy to think that your business needs a highly bespoke solution. After all, who knows your data better than you? But it’s a little more nuanced than that.
Our Secret to Choosing the Right Approach
Picking between Traditional and Minimalist approaches isn’t difficult, but can be a little unintuitive. When reflecting on which companies need a completely bespoke (Traditional) solution, versus one that can use a turnkey (Minimalist) solution, we were surprised to find that the difference came from a company’s revenue model.
Yes. Their revenue model. What models are we talking about?
Usage-based Revenue. Think of any product or service where you paid more for something the more you used it.
Ticket-based Revenue. Most commonly seen in Consumer goods and services, and Professional services. A ticket is a receipt you get from the Point-of-Sale or an Invoice issued after a project is complete.
Here’s the secret: If you’re usage-based, go Traditional. If you’re ticket-based, go Minimalist.
Why would the revenue model have any impact on the Analytics approach?
Think about the kinds of companies who have a usage based revenue model:
- Tech Companies: Per-user (or some other value metric) pricing; Purchasing subscriptions; cloud providers who charge based on usage; etc.
- Credit Cards: Buy more, higher credit limits, higher interest payments
- Insurance: Drive more, pay higher premiums
- Payment Providers: Stripe, Moneris, etc.
- Online Paid-Ads: Cost per click; Number of impressions
These companies’ revenues are tied to how much their products are used. The more they’re used, the more they make. So they’re highly interested in keeping you tied to their platform, increasing daily usage and reducing churn. They can’t know how much they’ve made until after the product has been consumed. They need to mine their massive amounts of usage data and route that through their billing and accounting systems to get a clear picture of their finances. This means their architecture tends to be highly complicated and very bespoke, and could look totally different from one company to the next even within the same industry.
Now look at companies that are more ticket based:
- Consumer Goods / Services (eg., apparel, retail, restaurants, etc.)
- Professional Services (construction, consulting, etc.)
You know up front the cost of a product or service. Whether you receive the ticket after payment (receipt) or before payment (invoice) is irrelevant. These are discrete, self-contained transactions unlike the former. And in all of these cases we’ve found that their internal data architecture looks 80% to 90% the same, even across industries.
This single distinction helps sort the vast majority of companies into one of two camps. If your revenue is usage-based, you’ll probably need a bespoke solution. Pick the Traditional Approach.
If your revenue is ticket-based, you’ll fare better with a turnkey solution. Pick the Minimalist Approach.
Another good question to ask is – who is the primary stakeholder and what are they trying to optimize? If you’re trying to maximize usage and minimize churn under the Traditional Approach, then the increase in top-line sales should more than offset the fixed cost of an in-house team. That’s because you have the scale to do it. On the other hand, while optimizing top-line is still important, it may not always be enough to offset fixed costs. In this case the stakeholder is the CFO who’s trying to maximize EBITDA, and would benefit from a Minimalist Approach.
Weird Exceptions to the Rules
We described the above revenue models as two “camps”, but they’re less discrete than that. Some companies don’t fit neatly into either, while others may fall somewhere in between. Here are some examples of companies that don’t quite fit the mold.
Mega Retailers. Think of huge companies like Wal Mart and Amazon. These businesses are obviously Ticket Based, but they have such a mass volume of data that centralization into one, 360-degree view of the business is impractical. So they have to hire an internal team to build a bespoke solutions for each line of business.
Public Sector. This one is a bit trickier. The public sector doesn’t generate “revenue” per se, and each ministry may be different. In this case the question to ask is if there’s any legislation that requires a ministry’s data be centralized and integrated. If so, choose the Minimalist approach.
Charities. Just like the Public Sector, Charities don’t really generate “revenue”, although they do raise funds through various means. However they may have mandated reporting requirements.
Real Estate. Another odd-duck is the Real Estate sector, who operate on very long time horizons, have massive returns, and may not feel the pressure to optimize EBITDA over a short time frame. But their revenues are still very project (ticket) oriented.
In all of these cases you have to dig a little deeper than just the revenue model and ask, “Is there a need or mandate to centralize the data?” The need may come from making decisions on a weekly basis. The mandate could be a piece of legislation. The second question to ask is, “Can this practically be done with as little customization as possible?”. For big businesses with legacy systems, the answer could be No.
But if there’s a strong need or mandate to centralize data, and it is a feasible endeavour, then the Minimalist Approach is a strong candidate.
Traditional vs. Minimalist Pros & Cons
There are pros and cons to each approach. And to be clear, even though TypeSift helps companies take the Minimalist approach, we’re not saying that one is better than the other. It truly depends on your business and its needs. Make the wrong choice and you’ll be trying to stuff a round peg into a square hole. The right choice and you’ll enjoy a higher return on investment, whether that means higher revenues (under the Traditional Approach) or higher EBITDA (under the Minimalist Approach).
| Traditional | Minimalist |
|---|---|
| Bespoke | Turnkey |
| Full-time, In-house team | Outsourced & Fractional |
| Great for Usage-based revenue models | Great for ticket-based revenue models |
| For companies who ARE data & software experts | For companies who are NOT data & software experts |
| More moving parts, higher maintenance cost, longer implementation times | Fewer moving parts, lower maintenance costs, shorter implementation times |
| Truly big data (Terrabytes and billions of log records) | Small to Medium data (Gigaybtes and < 1B records) |
| Customization over Configuration | Configuration over Customization |
| Optimize Daily active usage, customer lifetime value, lower churn. | Used to increase traffic (frequency of tickets) and/or ticket size (Average Order Value). |
Can you Combine the Two?
In a word: Yes. But you should have a very specific reason to do so. From our experience we’ve seen companies successfully combine these two approaches for one of three reasons:
- Avoid a Sunk Cost. You may have bought some of the tools and hired some of the people for a Traditional Approach, but want to make a pivot part way through. Rather than burn your past investment in the Traditional Approach, you can fit the Minimalist system into the established architecture. We do recommend striving to go totally Minimalist in the long term so that you can get the full benefit of it.
- Get the most out of your Data Scientists. Use a Minimalist Approach for the core RP&A but then hire skilled Data Scientists and Software Engineers to build AI & ML solutions on top of that. In this case, the Data Scientists don’t do work they don’t like to do (ie. reporting & data cleaning) but also benefit from starting on the same foundation of cleaned and consolidated data as everyone else. The Minimalist approach helps you get the most out of your data scientists.
- Split Revenue Model. If your business makes some money from usage and some from tickets, then you may split between the Minimalist and Traditional approaches, depending on what your revenue split between Usage and Tickets looks like. For example, you may take a Minimalist approach just for Finance, and a Traditional approach for everyone else. Or you may do the opposite, with everyone going Minimalist except one department (like Product or Operations). In this case, the Minimalist & Traditional approaches are fairly self-contained.
Get our help
Are you considering a Reporting, Planning & Analytics implementation? Want some help deciding on a strategy? Then get in touch with us. We’ll help you decide if a Traditional vs. Minimalist approach is right for you. If it’s Traditional, we’ll recommend vendors we trust. If it’s Minimalist, we’ll show you how TypeSift can help.
TypeSift is a Data Engineering & Design Minimalism Firm. Our expertise is decluttering information and solving problems in your data that are holding back your growth. We build software that corrals data and invokes ingenuity with the fewest moving parts.
