Statistics about Robotic Process Automation (RPA) and AI have fueled a very heated debate … a debate that demonstrates how to use data in B2B marketing.
Will robots take your job?
Or will they deliver sky-high ROI for your business?
In this article, I’ll use this debate to explain:
- Why some statistics should be taken with a heaping tablespoon of salt
- How to uncover the facts behind the data
- How data is used in B2B marketing
- And how you can use data to fuel your own marketing agenda
Below, we’ll dig into this data and explain why we should take it with a grain (or a heaping tablespoon) of salt.
Also, we’ll look at how data is used as marketing ammunition, how to dig behind the data to make informed decisions, and how data can be used to fuel your own marketing program.
Before we begin, though … what exactly is RPA? And how is it different from AI?
Quick Definitions: RPA vs. AI
RPA refers to software robots that are programmed by humans to perform specific tasks inside the workplace.
AI, on the other hand, can recognize patterns in data and “learn” on its own. It is not pre-programmed.
When it comes to automation, the two fields do overlap, as we’ll see in the data below.
What Robotic Process Automation (RPA) and AI Statistics Say About Jobs
There is a wealth of data about job automation.
Let’s look at a few trends in AI and RPA statistics – ROI, automation industry growth, and job loss.
Automation delivers massive ROI:
- Google used DeepMind AI to cut data center cooling costs by 40%
- Telefonica 02 deployed over 160 RPA “robots” that delivered a 3-year ROI of 650-800% … and that was back in 2015
- A report by Everest Group found that top performing companies earned averages of 78% ROI from RPA, while other enterprises earned 49%
- The Robotic Workforce claims that when correctly deployed, RPA software runs 90% faster than humans
- Cost savings can range from 10-50% according to the complexity of the process, says TATA Consultancy Services
Futurists predict that RPA and AI will explode in the coming years:
- Accenture says AI could lead to a $14 trillion increase by 2035
- Grand View Research projects AI to grow from almost $6 billion in 2016 to $36 billion in 2025
- Transparency Market Research predicts RPA will climb with a CAGR of 60.5% through 2020
- The RPA industry could will be worth $3.11 billion by 2025, says Grand View Research
- Acumen claims that figure will reach $4.1 billion by 2026
But a consequence of this explosion … software will take your job:
- The OECD predicts that 14% of jobs across 32 countries have a 70% chance of automation
- A report by Ball State University predicts that half of low-skilled US jobs are at risk of being automated away
- A well-known Oxford study predicted that 47% of jobs are at risk from automation within the next 20 years
- Brookings predicts that job losses in the US will be around 36 million by 2030
- In a study by Pew Research, 65% of respondents claimed that in the next 50 years, robots will do much of the work that is currently being done by humans
- Gartner predicts that smart robots will take over a third of jobs by 2025
As we can see from these latter predictions, there is a clear trend towards the dire.
The solutions advocated typically focus on re-skilling.
Some include political solutions, such as government stipends or universal basic income.
If You Rely on Data to Inform Decisions, Dig Deeper
The above statistics reveal interesting patterns:
- “Doom and gloom” predictions mostly come from private research firms, government groups, individuals, and educational institutions
- ROI-focused research mostly comes from private companies and research firms
In other words, data is being used for marketing purposes, plain and simple.
There’s always a motive behind the research.
And there’s nothing wrong with this.
Data is a tool. It should be used as such.
It’s useful for promoting one’s own agenda, whether that agenda is political, profitable, or otherwise.
In marketing, it’s fuel. Data is usually chosen to trigger an emotion, make a logical point, or both.
What’s important is to remember that data is not fact or truth … especially when that data is just a future forecast.
In some cases, those projections turn out to be spectacularly wrong.
Gartner, for instance, predicted that by 2018, 3 million people would be working under robot bosses. This type of prediction has earned searing critiques from some, most notably the tech research firm Horses for Sources.
Back when social media was still fresh, marketers loudly proclaimed the end of traditional advertising. Mass marketing and paid ads would give way to permission-based marketing, signalling an end to paid advertising. Forrester, Ad Age, and the Content Marketing Institute all joined in the fray.
Of course, we all saw how that turned out … advertising runs the internet.
I’m pointing this out to demonstrate that, as ad contrarian Bob Hoffman wrote, “The great thing about the future is that it can’t be fact-checked.”
Why You Must Take All Data with a Heap of Salt
Whenever we do research and use data to inform decisions, we must remember to examine:
- Data sources
- Statistical methods
- The motives behind the research
This last point — motivations — is key.
In some cases, research reports and studies are relatively unbiased.
In others, it’s not:
- Research firms sell research, so their statistics — present or future — are designed to sell their research and consulting services
- A white paper employs a survey designed to acquire data that supports its marketing agenda
- Governments have political agendas
I’m not saying this is bad.
On the contrary. It’s just marketing.
It’s what I do when I perform research and copywriting for my clients.
And, as you can probably guess, it’s what I’m doing in this very article.
Conclusion: Data Fuels the Sales Argument
Some marketing sells hard, some sells soft, and some doesn’t sell at all.
White papers, for instance, are not supposed to sell hard. That’s the data’s job.
That’s not to say that the soft or hard sell have no place in B2B marketing. They do.
However, when it comes to data, the sell is soft — sometimes so soft that it’s easy to forget the purpose behind the data.
Ultimately, whenever we evaluate data, we must dig deeper if we want to uncover the facts … especially if we are using that data to inform our decisions.
