Over the last few weeks I have laid out some key concepts of mind+machine analytics, so let’s now take a look at some trends in the world around mind+machine analytics, in particular the nature of the relationship between the service provider and client.
For most things in life, success means getting the right outcome. If the result is not what you expect, you probably won’t be satisfied. Why should analytics use cases be any different?
Decision makers want their invested resources to produce insights, not just gather a collection of data scientists, big data tools and AI algorithms. And yet, 99% of all analytics delivered still relies on input-based pricing, where the input is defined regardless of output.
Input might be defined by salary or hourly rate, whereas output is defined as a product delivered—for example, a qualified early-stage sales opportunity, a pitch book for investments, or a patent landscape. In mind+machine analytics the direction of travel seems to be towards output-based models, where the ‘units’ and the corresponding prices can be clearly defined.
Where does this trend come from? Clearly, cloud-based offerings in various domains have created demand for such models in analytics as well. For example, Microsoft Office is now sold on a pay-as-you-go (PAYG) model. It’s logical to apply the same model to analytics as well.
What are the drivers for output-based, PAYG models in cloud or enterprise mode, and what are the benefits for the customers?
- Low upfront costs and investments. Getting started is quick and easy. Most companies have threshold levels for investments that need to be approved by a central investment committee, which can take ages. PAYG helps to transform capital expenditures into normal operating costs.
- Improved transparency. Cost allocation becomes much more immediate and can be linked to the user in a far more transparent way, avoiding complex and non-transparent cost allocations altogether.
- Reduced risks. Due to the smaller commitments, the risks involved are much smaller as well.
- Increased agility. Being agile is becoming increasingly important. When companies use agile development methodologies for new products (e.g., 3-week sprints for software development), PAYG can make a huge difference. A new solution can be tried out and launched within a few days.
- Plug-and-play possibilities. When you combine PAYG with cloud solutions, the results are close to plug-and-play solutions. Such models need far less solution engineering than the current enterprise models.
- Always up-to-date. PAYG allows instant upgrades, as by definition, the contract period is so short that it allows instant renewal.
Of course, there is also the vendor’s perspective, be it an internal ‘vendor’ like a central data analytics team or an external vendor like InsightBee. There are advantages and disadvantages.
Let’s analyze the advantages first:
- Huge addressable market. A cloud-based PAYG product can be sold to everyone everywhere. This vastly extends your market.
- Attackers love it. Where sleepy incumbents are slow to adapt, the attackers are having field days. Companies like Hubspot in CRM and marketing or players in human resources, such as ZOHO, Workable or SnapHRM, are growing aggressively in a field formerly dominated by the heavyweight ERP system providers.
- Shorter sales cycles and lower cost of sales. Services provided in the classical enterprise model have extremely long sales cycles, from a few months to several years. For output-based models, the sales cycles and the costs of sales are much lower.
- Specialization. If there is enough scale in a certain use case, the potential of mind+machine can flourish to its fullest degree. The provider of the use case can come up with best-practice automations that improve the product significantly—and far beyond what captives would be able to do.
Let’s not forget to consider the disadvantages:
- Volume swings. An advantage for the customer, this can turn into a planning nightmare for the provider, especially in a mind+machine model.
- Lower stickiness and increased, global competition. Very short contract or subscription periods reduce the stickiness of the client relationship—and fully variable PAYG models even more so!
This model has worked extremely well for software providers thus far. In the mind+machine market, the first few models are becoming available now: InsightBee for Sales Intelligence or the Banker’s Studio for investment banking pitch books.
Of course, not all services will become output-based, but the philosophy of analytics use cases forces everyone to focus on the output rather than the input. With the predicted explosion of analytics use cases, output-based thinking will be essential if you want to keep costs from going through the roof.
In my new book, Mind+Machine: A Decision Model for Optimizing and Implementing Analytics, I provide more detail about how engagement models are changing and other trends that will affect data analytics in the near future.