PwC ups the M&A ante with cloud-based Decision Analytics Platform

In the first three quarters of 2020 there were 292 deals completed in Australia, with a deal value in excess of US$20bn. Across the wider Asia Pacific region there were 2,406 deals, with a value in excess of US$490bn1 . Mergers and Acquisitions is big business. Very big business, pandemic notwithstanding.

After identifying a shortlist of potential buyers, there is typically a period of weeks when the parties – whether buyer or seller – complete their due diligence, exploring financial and commercial track records. It’s not a lot of time when there are often many millions of dollars and corporate reputations at stake.

PwC is a leading financial advisor to both sellers and buyers; it has used the platform on more than 150 transactions in 8 countries since its launch earlier this year and expects to hit 400 projects by June 2021.

PwC supports its clients through the full deal cycle, from preparing a business for sale/identification of a target, through negotiation and diligence, and then finally on agreeing a deal. After a successful transfer of ownership PwC can also help the new owners maximise the value from the business they have just bought.

Depending which side of the transaction PwC is supporting (buyside or sellside), it may use data from lots of sources, including data from the company themselves (e.g. from their ERP/Sales systems), 3rd party data (e.g. Market studies) and public datasets (e.g. Government/demographic data). After the deal has completed, it can explore the commercials of a business in even greater depth, analysing customer purchase patterns, site by site performance and getting to grips with customer churn.

Five years ago PwC tackled all of this manually, largely with the support of Excel spreadsheets. More recently however it has utilised a different toolset to help streamline the process and has now deployed the proprietary Decision Analytics Platform, based on Microsoft Azure.

Charlie Pickett, PwC Director and leader of this program of work, says that while the platform was originally developed to support the yes or no decisions of M&A, it is evolving to support buyers as they navigate ongoing business decisions such as ‘Should we sell to this customer? Do we develop this new product? Do we enter a particular market?’

According to Pickett; “We are helping our clients use data to make better data-informed decisions.”

In the case of the recent My Muscle Chef purchase by Quadrant, the platform was used to ingest historic financial information from multiple sources covering both the online direct to consumer as well as their wholesale business, they also blended in demographic data to understand their customers better.

That information was then made available to multiple teams in PwC who were working on the transaction; “Analytics, financial diligence, tax, a broad set of people, and we could all work with the same data on the same platform in a collaborative way,” says Pickett.

“The killer use-case and benefit is having the same, single source of truth, the consistent platform experience, and the ability to give clients access to it as well. For that example, we gave Quadrant access to the data and the platform that sits underneath it, so they could actually pull out their own analysis. Given the time pressures we were under, this was invaluable”

We have a similar approach across all our projects, says Pickett; “Every transaction goes onto the platform and it’s managed in a consistent way. We can give access to bidders. We can give access to advisors. We can give access to our clients, all in a managed and secure way.”

Azure switch leverages broader ecosystem

PwC has been refining the platform over the last few years, consolidating the tech stack from earlier versions hosted across multiple cloud providers, to the current Microsoft Azure based solution. Pickett says that the breadth of products and solutions available within the Microsoft ecosystem made it particularly attractive.

“A lot of the tools that we’re using, like storage, you can get from any cloud provider, they’re commoditised. Even tools like Databricks are available off the shelf from multiple vendors. We still saw the benefits in consolidating the platform with one cloud provider.

“We are heavy users of Analysis Services as it gives our users the choice of which reporting/visualisation tool they use. We can support Excel as well as PowerBI and other visualisation tools with no extra work. If you want to integrate Analysis Services, alongside the rest of the data stack (Microsoft SQL, Databricks, Data Lake Store) only Microsoft ticked that box. Similarly, the ability to leverage Azure Active Directory to manage authentication and access was important given the strict security concerns involved with confidential M&A data”.

Pickett says; “There is an ecosystem of tools and technologies that talk to each other really nicely. We can develop the core application with the features we want to centralise, but other teams can also develop modules and integrations that sit alongside it, all in the same controlled Microsoft environment…we don’t have to worry about the integration. We just say, ‘Look, this is what we’ve got, this is what you’ve got,’ and the two bolt together really nicely because all the integration is handled by Microsoft.”

PwC has leveraged the Azure stack broadly, this includes the Data stack (Data Lake Store for data files, Databricks to handle data transformation work, Azure SQL to store the transformed data and Azure Analysis Services to layer on analytical calculations). PwC also globally distribute the app using Traffic Manager, CDN, App Services and CosmosDB to ensure a consistent and secure experience wherever the team is based.

Microsoft’s global reach and security track record were also important as the tool is already being used a number of PwC firms.

Access to Azure Multi Factor Authentication has also ensured PwC has been able to provide clients with secure access to the platform, confident that they will be able to access data relevant to their needs.

While Pickett and his team have led the program of work, they have been supported by third party developers and Microsoft specialists from around the world, particularly with the use of cutting-edge technology such as some of Microsoft’s recently announced serverless database features in both SQL and Cosmos DB.

He stresses the importance of a strengthened DevOps culture and capability. “Moving into the DevOps world means it’s a ‘button click’ to deploy every element of our platform. And we can put out tested code that we’re happy with, with the proviso that we can just roll it back if there are any bugs, because every single version is controlled, and it only takes two or three minutes to do the deployment. This means we can put out a lot more releases, meaning more features for our users, while still maintaining quality.

“That’s been transformational for us. I would advise any of our clients to do the same thing. If you’re going to run anything software related, get that DevOps pipeline working…it gives you the benefits of rigor and control with the flexibility of a shortened release cycle.”

Often the analysis that PwC is able to perform across the now comprehensive data collection reveals operational details that were previously opaque to the company in question. Although many may have had good analytics themselves – they haven’t always had the ability to query the sort of consolidated and comprehensive data collection that PwC puts together, bringing together financial data, sales data, and production data for example.

That comprehensive data collection has additional value after an M&A is complete, says Pickett. “We’re still working with a number of clients post-transaction, with examples like cashflow management, because as you know, Covid has affected trading for of a number of businesses.

“Using that same platform, but a slightly different focus. Instead of, ‘Do we buy the business, and how much do we pay?’ it’s, ‘How do we manage our cashflow? How much is coming in and out? How much might we have to top up, cash wise?’

“That’s the ideal. The transaction itself is just another milestone in the deal cycle, we should be continuing to work with clients whatever stage they’re at” says Pickett.

AI to identify best practice

He is now eager to explore how it might be possible to scale the deployment AI and machine learning on M&A transactions. Pickett is hopeful this would provide valuable business insights.

Pickett is not convinced that AI and machine learning will replace human advisors any time soon on actual M&A decisions – but he does believe that the technology could be useful in identifying the best way to operate once a deal is complete.

“I think using AI and ML to automatically decide deal-making is very difficult. And to be honest, if you do that, it quickly becomes commoditised. Everyone’s got the same information, at which point everyone’s back on a level playing field. That’s probably not the big push for us. I think the big push for us is using a client’s own data blended with third party data and our analytical capabilities to generate better insight and better data-informed decisions.

“We worked with a quick service restaurant group and pulled out the sales for each location that they had nationally. Then, we blended in the information about the shopping centres that they were in. How many stores were there? What other stores are there? How many parking spaces? What are the opening hours?

“Then, we also used publicly available information – demographics – median incomes, population statistics, all that kind of good stuff and blended it all together into a machine learning model, with the target of identifying which locations are inherently good and which are bad, and (importantly) to tell us which of these different factors (competitive profile, demographics etc) are important to determining that good or bad site.

“What we found, for example, was in one quick service restaurant group, they did well in areas that were nearby particular supermarket stores, and which had middle of the road income levels” – not rich, not poor – “this middle of the road was very good.”

For potential acquirers in M&A which may be new to a particular business area, that sort of data-backed advice is priceless and helps liberate the full value of the deal.