One thing I did is to look at what entities we find there. What concepts and ways to carve and define strategy? There are some obvious parts there, which are, of course, strategic goals and strategic assumptions which allow defining and test-driving strategy. I looked at that and then started to consider how can we put more structure into that.
For instance, a strategic goal to increase the market share in China by 10% in 2023 could be defined in a structured format. Market share in China is like an entity in an entity relationship model. That's basically what I did.
In addition, I also have a process model. What are the tasks like target search and processing the long and shortlist? So I defined the different tasks, and then I looked at the information or the data that are used by the different tasks and came up with this model that has structure and relationship between the different parts that make up a strategy.
This could also be used for defining consistency and completeness and provide some red tape to say, what does this strategic goal relate to when I can look it up by following the red tape.
Because there are a number of issues you face today. Strategy is separate from the usual business operations. Some extracts of data are used, and they are massaged and altered during the very manual strategy process. I wanted to change that. I wanted to have the same underlying data model for strategy so that I could also define a relationship between the data I need in strategy work.
That might be generated from the data I have in market information in databases out there, with company data, and so on. So it would make automation in the strategy process a lot easier by having the data structures and the tasks for the M&A strategy as well
The strategic entities are high-level concepts like markets, customers, solutions, suppliers, and partners often used to define certain strategic goals. For instance, we may want more partners or double the revenue from indirect sales via partners. These concepts - revenues, partners, etc. - are the objects in the statements.
That's what I took and made it part of the model. As soon as you have these objects, you can define the goals as I just explained. Strategic assumptions are basically assumptions about any of these strategic objects.
For example, the market will grow with a certain compound annual growth rate or market access will be possible for a company in a restricted market, among other assumptions that you could take. These form the hopefully solid foundation for a strategic plan and for the strategic goals.
The metamodel in my book defines the underlying spider web of objects and relationships. For instance, you have a strategic goal, it has to relate to a strategic entity. This structure can also be used to conduct consistency checking, to walk along the relationships to see if things are correct, consistent, even complete. This would hopefully make the strategy better.
One of the motivations for writing this book was to move from a vague approach to a structured data model, which is always interesting. One of the key goals was to transition from a PowerPoint filled with strategic goals, assumptions and plans to a structured data model of strategic assumptions plans. That was one of the main drivers for working on an overall data model for the M&A strategy phase.
Take for example, the strategic entity go-to-market, which contains all parts of a go-to-market endeavor. This includes the people working in go-to-market such as the salespeople, the pre-sales people, and also the different markets they operate in. All of this is encapsulated in the strategic entity go-to-market. It also contains the application systems that are used in go-to-market in different ways, like a CM system and others.
This provides a comprehensive and detailed picture of what go-to-market entails, but it's encapsulated in this strategic entity go-to-market. Hence, you have both a high-level model as well as a very detailed model if you want to drill into details.
A general theme in go-to-market, particularly when based on SAP solutions, is the dramatic decrease in the level of guesswork. There are excellent ways to leverage existing information about what's happening in go-to-market, about customers, their potential needs, and what we could sell to them.
This data can be used to predict numbers with machine learning models more effectively than by asking your regional sales lead how much more revenue they can generate if we buy a particular company. There are already ways to use what I would call advanced analytics tools to predict the revenue that could be generated by acquiring a certain company.
The model itself can be used to set up and also change strategic plans accordingly. Over the course of the M&A process, there's a lot of information coming in. The amount of information increases over time. You start with a very early draft of the business case, then you have more information, you acquire the company, you have more information, and so forth.
It's an ongoing and evolving plan from an M&A strategy to integration and through integration, with many changes along the way. The model itself would allow for this ongoing and frequent change. However, it's not yet implemented in a tool or an application system that we could really prove that. That's the next endeavor to collaborate on.
Breaking it down into details to complete the strategy
Just imagine you have a set of three strategic entities: customers, suppliers, and solutions. You decide to create an M&A strategy, which might impact all these three strategic entities. One way to ensure completeness is to say, for instance, we want to enter a new market. We want to have new solutions for that specific market and have strategic goals for that.
But if you don't have a strategic goal for customers, then this is evidently an incomplete strategy. After all, one of the ways you enter markets is by acquiring new customers in that specific market. So there is an inconsistency there. By reducing the vague elements, you have methods to easily check this kind of completeness and consistency within the model.
It's a structured way of executing the tasks we already undertake. We have very detailed discussions in strategy. We put a high effort into detailed due diligence, with careful plans outlining what we want to do and what we want to avoid in due diligence and post-merger integration. It's more about structuring what we do already to avoid overkill.
Consider the abstraction layer of strategic entities. You might not care about what's inside the strategic entity 'market'. It may contain 20-30 data objects, but on the high level, you're not concerned with that. You simply declare a strategic goal, such as raising the market share. Later, if you want to execute on that strategic goal, then you can use the more detailed level.
To decipher what it means to raise the market share, you might need to increase advertising or hire additional salespeople, or take other appropriate measures to reach the goal. These could then be cascaded or detailed out on a more detailed level. So, you have this manageable level with strategic entities to define strategic goals.
Strategic Fit Analysis
That's a crucial aspect of target search as well as long and shortlist processing. As a provider of products and solutions, naturally, there's a focus on products and solutions to determine the fit. However, that's not enough. Let me provide a simple example. Suppose you're a manufacturing company with the goal of being a high-quality, high-price provider.
If you acquire another company that is a price leader, you can't merely say the product fits well. There might be tasks to undertake because that company has an entirely different pricing and quality strategy. So you have to examine all the strategic aspects and compare the companies to arrive at a proper statement about the strategic fit.
In such a scenario, you would need to decide whether the target should continue the price leader strategy or transition into a high-quality, high-price company.
Consider the example of target search. I've examined 17 different tools and the data they use, and data is, of course, a significant topic. One end of the spectrum is using all the data available on the internet. At the other end, solutions like Embryonic from EY use all the prominent, costly company databases out there. Both approaches have their merits and drawbacks. If you use several databases, you need to curate because some data, like revenue data, might be better from PitchBook than from Crunchbase and vice versa for technologies.
On the other hand, extracting information and making sense of it from the internet, as COTAs does, is yet another approach. Both methods have their advantages and disadvantages. There's a huge demand for data and a significant need to create appropriate strategic assumptions and strategic goals that you want to ensure you can execute.
So, there's a high requirement for data access as well as a method to solidify and truly make the strategy consistent and complete, so you can also execute it. What's the best data set? It depends. Many providers claim to have global company data sets.
Then there are numerous companies who say they are local, they are in Scandinavia, and they know all the Scandinavian companies, startups far better than the global database because they're not present in the countries. Others claim to have all the Chinese startups covered better than others. It really depends on the exact scope of companies, markets that you're searching in.
Starting with one of the well-known large company databases is always a good idea, but you might have to look to other vendors as well for very specific industries or countries to have the best data set.
Cascading of Strategy
Cascading of strategy is a recurring issue in all strategy work. For instance, the CEO announces, 'Let's become a high growth company and extend our market share in the US by 25% within two years.' Everyone knows what the strategic goal is, but what does it mean for the sales representative in Cincinnati? What is his role in achieving this goal?
To ensure the sales representative in Cincinnati takes the correct actions to contribute to the overall high-level strategic goals, you need to cascade the strategy. He needs to know what it means at his level of the company hierarchy. For this reason, strategic goals can be broken down to make them more easily digestible and executable by individuals in any function within the company.
In my model, there's a direct connection from these high-level strategic goals - always connected to these high-level entities - to a more detailed inside view of the strategic entity, which allows them to break down their strategic goals into more detailed goals on a more granular level of the organization and data.
For the Cincinnati representative, this might mean that his focus should be on high-priced items, or specific customers who urgently need to transform their business. Whatever that means, it should ultimately contribute to the high-level strategic goal of increasing market share.
This would cascade all the way to the task to execute your diligence and integration ultimately. The good thing is that you cannot only cascade, but you also have the red tape to go back and show what is exactly your part in the overall game to reach.
And that's a good thing that you can go both ways. It's also a motivational thing for the Cincinnati guy to know what his share is in this overall goal and that he has also proven to be part of the overall aspiration of the company.
Changing the strategy
The planning process always involves iterations. At certain points, it's important to understand that changes may occur. So, you should have some buffers in your planning, in your budgets, and in your revenue forecasting that you can adjust as more information becomes available. That's the overall approach, which is always iterative.
There are specific moments where decision points occur, like going to the executive board for approval to enter into negotiations and the due diligence process. At these points, we take a snapshot of the plans, providing a caveat that this is only the current snapshot, and changes may occur. There will be an update in the final decision meeting.
It's crucial to make everyone aware that, at any given moment, the plans they're looking at are not final. There could be changes needed for us to be successful at a later point in time.
Learning from the target company
The importance of learning from the target company gains greater significance as you gain more experience in acquiring and integrating companies. A crucial question to ask is, why exactly did we buy this target?
We've had cases where we acquired a target because it operated in a new market unfamiliar to our company, or it ran a business model we hadn't previously implemented. In these cases, it's clear we purchased the company because they knew how to run that specific business, and it wasn't our place to tell them how to operate because we simply didn't have that knowledge. We had to learn from them. This mindset and its subsequent consequences are vital elements of success.
If you tried to quickly and fully integrate a company whose business, sales, and operations model you as the acquirer aren't familiar with, you risk damaging the company, leading to an unsuccessful integration. As an example, if SAP acquires a company with a business model it can't immediately execute, we would avoid a quick or complete integration until SAP, as the acquirer, has learned how to run that business model successfully from the target company.
Let's consider simple things like the company's policy regarding developers' computers. One company might always opt for the highest-priced models for their developers, whereas SAP typically chooses the second highest. This difference can potentially frustrate the developers on day one, prompting a reconsideration. Do you want to upset all the developers from the start, or perhaps exceptionally go for more expensive computers to welcome them and not reduce their capacity? These are changes you must consider.
This example is simple, but overall, we aim to predict as many potential obstacles as we can. We also think, plan, and budget for the mitigations needed to overcome these challenges in post-merger integration.
That's a really good example. It shows the necessity of being able to adapt and change plans quickly, and the importance of fostering a culture open to such flexibility.
Absolutely. This work is ongoing. Change is always difficult to enact in large company environments. That's why we support the acquired businesses for a significant portion of time in post-merger integration until we believe they're ready to continue in a regular day-to-day environment, separate from the protective umbrella of post-merger integration.
Quantifying culture in M&A
Yes, we briefly discussed this earlier: culture can be quantifiable to some extent, or at least incorporated into a data model that describes strategy. You can easily observe this in any surveys done with the acquired team, for example, to gauge their wellbeing, motivation, or any issues during integration. In these surveys, you find elements like attitudes, norms, beliefs, and other factors that should be considered when addressing culture.
However, not everything about culture can be quantified. Much of the cultural exchange between the target and the acquirer is based on experiences. We have a team focused solely on the experience of the acquired employees. Many things, such as role models and good examples, can be extremely helpful in making the acquired employees feel at home with the acquirer.
Impact of new emerging technologies in M&A
Let's start with the current market and how we might advance into the future. The predominant vendors in the M&A tools market today are the data room vendors, and they are expanding their scope beyond data room functionality and due diligence into post-merger integration (PMI) and earlier phases. From my perspective, these will be the established platforms for M&A automations.
However, I also believe that they won't be the sole providers of innovation. They will need to integrate other very focused and innovative solutions into their offerings. The best way to do this today is by providing open APIs to allow other solutions to plug in. For example, you might have a data room but want to use highly skilled automatic contract analysis from another vendor. You would then just plug it in to the data solution to take advantage of it.
Automation in M&A
Different platform vendors, including the data room vendors, are starting to provide machine learning based functionality in various forms. However, what I see in other industries, and miss here, is machine learning enabled analytics.
For instance, information such as growth rates and revenues could be automatically collected and perhaps combined with the acquirer's information to make predictions on potential revenue when our sales force sells the target solution.
We already see glimpses of this. For example, a company called Modernizer offers automatic sales predictions. It helps determine which solution would give you the highest likelihood to strike a deal when approaching a new customer.
We're currently working on using similar technology not just to predict sales to one specific customer, but to determine the overall revenue potential for the acquired product for the acquirer. So we're moving from assumptions to predictions.
Indeed, you can make better predictions with more formalized assumptions. One idea behind the book is that having structured the domain of M&A strategy makes it easier to build solutions and pinpoint the data needed in certain situations, and then work on top of that.
During diligence, we know that there may be thousands of decisions that are not typical day-to-day choices that must be made during an M&A process. As soon as you have proper data, this could be used to augment the information for decision making.
In different contexts, there are solutions already in place. For instance, if you run a car tire company and are in the process of ordering tires for the winter season, there are analytical solutions, decision support tools that provide you with a proposed number of tires to order based on new car sales and weather predictions.
That's exactly the kind of technology we should be using in the M&A process to make better decisions. As soon as you know what decision to make, there might be machine learning based algorithms that gather information and make proposals to help you make better decisions than before. This is because they provide augmented pieces of information that aid in making improved decisions.
The late evolution of M&A
You're absolutely right with your point, but there's also an additional aspect. People in corporate development are not typically technology enthusiasts. They may be proficient in financial models and the like, but not necessarily in understanding the latest technology.
We've discussed the use case for predictions. This falls in line with what Kahneman termed, 'What you see is all there is.' If you're not technologically inclined, you may not request a certain technology be used in the M&A process simply because you're unaware of it. You only know what you know.
This is why I included a section in the book that explains different technologies and how we could leverage them in the M&A process. We've discussed predictions, automatic summarizations, among other things. Once you're aware, you can create some pressure from the customer point of view and ask, 'Why don't we have this sophisticated machine learning tool in the M&A process as well?'
Then, we'll see a much higher degree of innovation in the M&A tool landscape.
M&A needs better tech leaders to drive the aspiration to innovate. Then, we can have better M&A outcomes and a superior people experience for everyone. Absolutely.