Pre-amble.? A couple of weeks ago I attended the IBM Think conference in San Francisco. It was the first time for the conference in San Francisco and, taking into account the learning curve, I like the change. ?It got me thinking about process vendors and their relationships with the fast-growing RPA segment.
- Pega bought OpenSpan a few years ago, with a strategy to tightly integrate OpenSpan into Pega?s offering. ?This happened before RPA vendors? valuations got truly out of hand for M&A purposes.
- Appian partnered with and maybe white labels BluePrism
- A couple of other BPM vendors have partnered with UIPath (including Bonitasoft and Camunda)
Separately, many of the big consultancies and BPO shops partnered up with one or the other of these RPA vendors - usually Automation Anywhere, BluePrism, or UIPath, though there are many others.
In the last 2 years, RPA has evolved from being ?something of interest? to all the software process vendors, into something of critical importance to these same businesses? plans and strategies.
What about IBM? IBM has been a leader in business process, content management, data capture, and other important enterprise technologies, but historically IBM hasn't played in the RPA market ?- until recently.
This isn't unusual - most of the business process vendors were late to the game. Given that UIpath and AA raised megarounds in the last 12-18 months, it's hard to see (from the outside) how an IBM could justify buying an RPA vendor outright (and hard to see how the smaller BPM firms could buy any of the RPA firms of note). ?However, RPA is more important to IBM's clients and their strategies than ever - it can?t be overlooked or ignored. So what do you do if you're IBM?
Why not buy? ?I'll lay out our observations and conclusions based on public information available. ?First option, why doesn't IBM buy an RPA vendor?
- valuations are high, revenues are still (relatively) modest
- fresh investment rounds in each of these RPA companies make it too soon for them to divest
- IBM's clients by and large have RPA software already - possibly more than one RPA vendor. ?IBM services have a track record of delivering services on any RPA platform
In conclusion, buying an RPA vendor outright is expensive, has big risks, and runs the risk of alienating a swath of clients who have chosen a different RPA vendor.
Second option, why doesn't IBM write its own RPA tool?
- The simplest answer is, there are already great RPA products on the market, and it just makes sense to work with something that is already working for clients. Software companies often fall into the temptation of always writing their own solution, I'm actually impressed when they have avoided this pitfall and have partnered instead. ?Don?t reinvent the wheel (again).
- It is always more expensive to build a product than one expects, and it takes time to build it, during which you?re not addressing your market.
- Also: IBM would be swimming upstream to sell a new and unproven RPA tool to clients who already have one.
Third option: ?do nothing. We already discussed why this isn?t an option for IBM: part of IBM?s mission is to stay relevant to their clients, and having an RPA strategy today is part of being relevant.
So what is IBM doing instead? ?
IBM is taking a fourth option: if you can?t beat ?em, join ?em. ?IBM is partnering with RPA vendors, and embracing and extending RPA with IBM?s core strengths: content, data capture, process... and AI. IBM is announcing offerings that leverage those strengths in combination with RPA - and focusing on the value of those combinations - why?
- First, instead of replacing one RPA product with another, IBM is focused on increasing the value of any RPA product IBM clients are using, through improvements to IBM?s other offerings.
- Second, IBM can give all their clients with RPA a big hug. ?IBM has the benefit of not alienating clients who have chosen a particular RPA product - the big three independents are supported, and potentially any of them could be supported with a little TLC.
- Third, IBM is betting that they can learn how to embrace RPA in a big hug faster than RPA vendors can learn how to apply machine learning and AI to solve the problems IBM is tackling.
It's an interesting bet. ?I mentioned content. Specifically, data capture and content analysis is really a machine learning topic. Pulling structured data from unstructured documents, or from analog documents, is a huge feat of machine learning. ?But we just call it something innocuous like "content analysis" or OCR. (Thus the old adage that it's called "AI" until it works, and then it is called something like "Search" or "Rules" or "facial recognition" instead)
If IBM can parley its strengths into a fast-growing RPA market, it can potentially benefit both itself and RPA vendors, in addition to clients. It isn't an easy path, but it is a clever strategy for a market that suddenly exceeded everyone's expectations. The elephant can, apparently, dance with the nimble RPA "startups"!
At BP3, we can relate to this strategy. After all, what do services companies like ours do? ?We focus on how we can bring our talents to bear - around other companies? IP or capabilities - to deliver a more well-rounded approach. We?re looking to leverage our strengths, and also leverage third party software. Our own journey toward RPA started with understanding that RPA was becoming important to our clients, and we needed to understand why, and make it important to ourselves as well. ?This is *what we do*. I have to admit I found it interesting to catch a company the size of IBM employing much the same strategy, leveraging *their* unique capabilities on top of someone else?s core IP.
I guess the question now is, how will the other process vendors respond to RPA? So far they have responded primarily by partnering. Will consolidation make more sense? Will RPA firms be the more likely buyers? Or will RPA firms attempt to add process capabilities to their very capable task automation (RPA) products? Time will tell! Appreciate your thoughts and comments below!