M&A, Business Models and Ecosystems in the Software Industry

Karl´s blog

Posts tagged ML
Software strategy selection: is build, buy, partner sufficient or do we have to add open source to the game?

Strategy selection

The best innovation and growth strategy is to combine organic and inorganic growth. SAP has successfully applied organic innovation and growth resulting e.g. in SAP HANA, SAP S/4 HANA as well as inorganic innovation and growth via acquisitions like Qualtrics and Calliduscloud. 

Build, buy, partner

  For me, the most important distinction between build or buy is the window  of opportunity that you have.  In technology markets, there are frequent changes of market direction. If you’re lucky, you had started  your solution  in time to build something that is en vogue  right now. But if you’re not lucky, you need to acquire capabilities that the market needs today.  But is this the only option you have?

Opportunity and risk in building  and acquiring solutions

To be frank,    with the current state of technology due diligence on to be acquired companies there is no difference in risk to build or to buy.  When building products, you trust your developers to build something great. The a priori likelihood of success is 50%. Same likelihood applies for acquiring technology. In addition, acquired technology exists, has customers, success and failure history. So, what is the impact of this statement on build decisions? 

Build decisions

Build decisions are made based on anticipated market trends. So don´t be suprised when you find out that you made the wrong decision. It is perfectly natural to take wrong decisions. But how can you fix such a wrong decision? I have two proposals: The first one is to start massive marketing to convince customers and markets that what you built is the right thing. Tough. The second option is to buy your way into front and center of the market.  What are these the only options you have?

 Outsource your worries

 What we need to look at is in another alternative.  You could leverage an existing open source solution with a license that permits commercial use to jumpstart your building efforts.   And you build differentiating, proprietary technology on top.

 If the open source community behind that solution is being active enough, you will save massive effort for support and maintenance of the solution.  

It also makes financial  and strategic sense to spend your money wisely on functionality where you can differentiate your offering from the competitors’ offerings.

Why don´t you choose one of the following topics to continue:

 

How machine learning can help in digitalization of M&A processes!

Machine learning is everywhere - except in M&A processes. Let´s change that. Let us imagine the impact of machine learning in different steps of a typical M&A process. Let us start by sharing some of my ideas to trigger your imagination. I am convinced that the technologies needed to achieve this vision are in place today, they are just not being used in this context.

Early phases of the M&A process, shortlisting phases

Let´s say you have five companies in your shortlist. Machine learning can help finding and selecting potential targets e.g. by predicting which of the companies considered will be the unicorn, i.e. the most successful company in the list. Approaches for doing that exist, e.g. https://papers.ssrn.com/sol3/papers.cfm?abstract_id=3159123

Preparing the Letter of Intent

Based on past projects, machine learning can help to predict deal breakers, find missing or potentially wrong data in the financial valuation of the target and propose deal structure and clauses for the letter of intent based on the existing, available data about the target and the acquirer.

Due diligence

A vital part of the job to be done in due diligence is that you are looking for missing data, for deal breakers and risks in documents in the data room BUT you only have limited time and a huge data lake in the data room. So let us see how automation and machine learning could help us here.

Day one of due diligence: the data room is available. Day 2 of due diligence: Information about missing data, deal breakers and risks is already available.

How is that possible? Using automated document/contract analysis based on machine learning as well as data about deal breakers and historic projects, a machine learning application can provide this information. There is a huge value in this: you get more time in due diligence to work on missing data, for deal breakers and risks, so quality of due diligence results will massively increase.

No more reporting: During due diligence, digital assistants will automatically keep the lists of tasks, risks, issues and results, will create automatic reporting from that and propose next steps.

Merger integration

Results from the due diligence are automatically distributed digitally to all integration team members. Machine learning based digital assistants propose the integration plan, the integration timeline and which next steps should be taken. They analyze due diligence data and propose the set of data that should be doublechecked and validated. They validate that data by extracting information from the target´s ERP systems automatically and present deviations in digital dashboards and propose next steps.

Learning assistants analyze the learning needed by the involved integration managers based on their CV and proposes digital learning lessons based on PMI2GO.

No more reporting: During due diligence, digital assistants will automatically keep the lists of tasks, risks, issues and results, will create automatic reporting from that and propose next steps.

Let us imagine the impossible - and make it work

The opportunities are massive but are not yet leveraged. I think the M&A community has to provide guidance to vendors to achieve a vision i call the Digital M&A Manifesto. Stay tuned for more details. Like this article to get more inspiration!

Digitalization of M&A: robots are boosting M&A process performance

While we are used to physical robots vacuuming our homes, software robots are not in widespread use yet. The term used for software robots is robotic process automation. (RPA)

What is RPA? 

RPA is defined as tools to build automation for everyday tasks and processes  on a computer screen using Software Robots.   This can start with a simple sequence of clicks on the screen that you can replay automatically. But RPA can also cover more complex workflows with decision points. RPA  tools usually contain a recorder that tracks  certain work sequences on your computer screen and can replay it this sequence later.

What is RPA combined with machine learning? 

Recording workflows with current RPA tools is a manual process. If combined with machine learning, a digital assistant will track your online work and will propose automation of routine processes you do every day. This will lead to a step by step increase of the level of automation in processes.

How does RPA help in M&A processes?

 It frees up time to focus on the really important topics instead of routine tasks  and sequences of clicks on a computer screen.