The Author: Anne Golombek
Anne Golombek is COO and Marketing Lead at minubo, the Commerce Intelligence Company. As an expert in data-driven commerce, she is one of the initiators of the Commerce Reporting Standard project.
On Thursday, November 8th, our third CRS workshop took place in the Project A offices in Berlin. With participants from companies like REWE vertical ZooRoyal, Sportspar, Shopmacher and of course our co-hosts from Project A, the Otto Group and minubo, we had productive discussions on our current topic: The Business Vitality Model (BVM). Here’s a summary of our main discussion points.
Three Levels of Application: Strategical, Tactical, Operational
Business decision-making can generally be split into three main levels: strategical, tactical and operational – and so can the application of supporting analytical models. Following this, the main question of our discussions was: At which level(s) can the Business Vitality Model best be applied? Let’s go through them and have a look at the key insights we gathered on the single subtopics that unfolded within this discussion.
The Strategical Level
As the model has been developed from a strategical use case, application on this level had already been worked out pretty well before the workshop (see e.g. our previous article). The key aspects summarized:
Business Health & Evaluation
As a normalized, integrated model, the BVM is a great way to evaluate business performance and look at the interdependencies of multiple parameters. With that (vs. only showing bottom-line development), it can deliver true transparency on the health of a business – and the functioning of the underlying model (see our example from our first article: bottom-line development looks okay, but the BVM shows an undeniable melt-down of the existing customer base – and with that, crucial business substance). Due to its standardization, the BVM does also offer a great way of benchmarking.
Forecasting, Sensitivity, Scenarios
On top of that, as stated in this article already, forecasting, planning validation (against own business history as well as benchmark data), sensitivity analysis and scenario analyses are proper use cases at the strategical level.
Communication & Cultural Change
As worked out in the discussions, the model also holds a great deal of value in terms of company communication and cultural change: As the model can be broken down for usage on all levels and areas of business, it is perfectly suited to communicate business development to the staff and break high-level goals down into tactical and operational goals for every department. On top of that, with the cohort view being both customer-centric and well explainable, it is a good tool to get traditional business controlling (that’s still the status quo in many businesses today) to adopt a more customer-focused approach. With the BVM as not only an analytical model, but also a mindset, this turnaround can be made.
Differentiation of Stakeholder Goals
Of course, the “strategical level” is usually composed from multiple kinds of stakeholders with multiple kinds of goals, so further differentiation is important for an effective usage of the BVM. Example: Whereas company management might be focused on data-driven planning and planning validation, their investors (or the company holding) might be mostly interested in the validation of the general business case, investment logics and market positioning.
The Tactical Level
On the tactical level (as e.g. department leads), model application moves to another level of detail. It’s not so much about overall business health and development anymore, but about the development of specific areas and drivers.
An important aspect of discussion was the differentiation of business dimensions on the tactical level of decision-making – meaning to look at performance development from various business perspectives. In our discussions, the following dimensions were identified as the most important ones for the application of the BVM:
- Customer: The basic version of the model (see here) is already very customer-centric, of course, but the new vs. existing customer analysis is not the only way of using the BVM from a customer-focused perspective. In the workshop, we did also discuss more specific cases like comparing long-term buyers that in some business models might still originate from the catalogue age with a new generation of buyers that genuinely interacts online – e.g. in terms of margins or assortment preferences. This kind of segment and lifecycle/lifetime potential analysis can hold a great deal of value for decision-making on the tactical level.
- Product: A key question during the workshop has been, if with the BVM, companies might run the risk of losing sight of the assortment perspective on their business. Our conclusion: No! The BVM can also be adapted to product-centric cases like e.g. the question which entry products brought valuable customer cohorts. With insights like that, tactics like the allocation of discounts on certain products can be discussed to ensure initial buy-in of customers with a high likelihood of long-term value potential. Another example would be the differentiation of mono and mix buyers (category-wise) to manage promotions accordingly.
- Channel: Using the BVM from the perspective of marketing channels, there’s a lot of insight to gain on customer behaviour – especially when looking at the correlation of channels and customer segments (e.g. entry channels of certain customer types, channels that initiate first purchases, channels that deliver high returns or low margins…). With that, an alternative approach of looking at customer cohorts opens up: splitting them based on first contact date instead of first purchase date. We agreed that both methods can hold a lot of value for different analytical questions; basically, you can say: The more operational the use case, the more important does the first contact method get.
- Market: Also for analysis and monitoring of external factors can the BVM be leveraged. Here are some examples we discussed: Monitoring how far a channel has already been siphoned off (especially when a company grows faster than that channel); making the mobile shift transparent via a breakdown into device-based cohorts; monitoring the effect of Amazon’s advances into specific product categories by making altered customer behaviour and acquisition figures transparent or – the opposite case – uncovering the effects of commerce giants going bankrupt (like e.g. the Arcandor Group in 2011 that led to a boost of new customer acquisition figures in several other commerce companies) to be able to differentiate between expectable and “lucky” growth effects for further forecasting.
- Linking of single business dimensions: Finally, it can be quite valuable to link the single dimensional views – like e.g. in the case of monitoring entry categories for valuable customer segments or cohorts. Views like that encourage interdisciplinary thinking (e.g. purchasing considering which customers they are about to acquire with certain products) – resulting in a more holistic business approach.
Break-Down Into Monthly Cohorts
Starting at the tactical level, a break-down of the model’s annual cohorts into monthly cohorts should be considered for some use cases. For example in working out which customer segments/personas to focus on, in gathering insights on seasonality or in allocating marketing budget to single channels, development during the course of the year should definitely be looked at.
The Operational Level
On the operational level, we get to the most detailed business view, of course. Actually, to make best operational use of the BVM, the level of single transactions and customers would be needed – maybe even daily or at least weekly cohorts. And talking about operational usage in marketing: Wouldn’t that require the integration of attribution models as well? Ultimately, all of this results in one key question: Up to which point does the application of the BVM make sense? Up to where does it really help – and where does it get too complex? Actually, we did not answer this question conclusively yet, but this is how far we’ve come
Goal Derivation and Reporting
First of all, as already mentioned above, the BVM offers great value in developing business goals based on one holistic model. From the same database that provides the overall company goals, department goals as well as even more detailed operational goals e.g. for weekly achievements or single customer segments can be derived to ultimately result into one big picture. That does not only make goal development easier, but also enables better goal tracking and reporting as well as easier monitoring of the effects of single goals being under- or overachieved.
Common Communication Basis
Not only in terms of goal derivation and reporting, but also in a more general way, the BVM can serve as a common communication basis throughout the entire company. With a model like this anchored in the minds of all stakeholders, everyone speaks the same language, so communication between departments, between team members and superiors, between management and team leads as well as between management and investors gets easier. Not only thanks to the model itself, but also through the fact that a model like this requires solid definition of the underlying metrics. We did also cover the topic of metric definition during the workshop, but let me just refer to the definition of transaction metrics that our project already achieved last year – with that, you get a holistic view on key metrics that might be really helpful in the process of developing a consistent data model.
Use Cases Derived From the Tactical Level
In terms of specific use cases on the operational level, a lot can be derived from the ideas that are relevant on the tactical level as, of course, we’re regularly talking the same goals here – broken down into more operational views. Accordingly, the differentiation of new customer acquisition and existing customer reactivation on campaign level would be an example for a marketing use case; just like clustering buying frequencies or customer preferences in terms of e.g. communication medium – all of that to make better decisions on which marketing activities are worth an investment how and when. At this point, we’re also back at the point when considering cohorts based on first contact date instead of first purchase date can really make sense for many analytical questions.
Levels of Application: Conclusion
Based on these discussions, the plenum ended up with the conclusion that the focus of the BVM stays a strategical one – corresponding to where its formative idea originally came from. The main reason: The differentiated construction of the model that delivers so much value on the strategical level by bringing transparency into interdependencies between multiple business drivers just rapidly gains complexity the more operational it gets. Though, we were able to work out clear added value for the tactical level as well and are convinced that, implemented with care, the model can even deliver value for single aspects in a company’s operational work.
Outlook: Developing Effect Chains
At the end of the workshop, everyone agreed that the next logical step in working on the Business Vitality Model would be to develop concrete effect chains. For example: Start with the budget, derive campaign outcome, derive customer acquisition/reactivation, derive revenue, derive margins. Or: Start with customer acquisition/reactivation, derive revenue and needed budget. Or: Start with customer segment development, derive campaign planning, derive needed budget, derive revenue. There are multiple effect chains in every business – depending on the perspective you have, the premises you start with and the goal you want to achieve. To provide a solid starting basis for everyone who wants to leverage the BVM for their own business, we think about developing a couple of frequently needed ones in the near future.
Don’t miss out on this and our next exciting topics! Sign up for our project newsletter and we’ll keep you posted.