Microsoft Access took the world by storm in the 90s and, though a fantastic solution, many of us have spent the decades since cleaning up some of the messes made. As we enter largely new territory in building digital workforce capability, there are plenty of mistakes that we can avoid making this time – but only if we borrow learnings from the Access debacle to set ourselves up for automation success.  

Cast your mind back to November 1992, if you can. You might recall Bill Clinton’s decisive victory in the US election, or you may have more vague recollections of acid wash denim, tan leather jackets, and shoulder pads. Perhaps the more notable event happening though was the birth of Microsoft Access.


Whether you remember its launch onto the market or not, you’ve no doubt felt its ripples since. Microsoft Access made people believe that building software was easy and Access-based systems began popping up everywhere as the masses – many of whom didn’t really know what they were doing – used it to build things for their businesses.

To some extent this was a good thing – the solutions did solve pain points and some people developed amazing stuff using Access. However, it didn’t always have the intended impact and there has been an abundance of badly built systems that emerged as a result. Many of us have spent the decades since dealing with these, which leads me to my “poisoned chalice” analogy.

I was recently speaking to a former colleague about my favourite topic – robotic process automation (RPA) – and he mentioned that he can’t help but wonder if RPA is just that: a poisoned chalice. I took this to mean that despite looking good, and making a whole lot of promises, RPA could also prove to be a source of future problems. And I had to agree, but only to a point.

The benefits of increased sustainability, agility, and efficiency that automated digital workforces can deliver (and have delivered) for many organisations are massive – but they have to be done right to have a positive impact. Above all, we need to not mistake immediate benefits for the makings of a long-term solution.

The thinking about, and implementation of, RPA is still in its infancy – and while there are some trailblazers who are doing it well and successfully creating digital workforces that deliver positive ROI and help an organisation scale, some are struggling to get the results they need from automation.

However, society, business, and the way that we all adapt is cyclical. We may be in new territory now, but we can learn a lot from what we’ve done too. Fast-tracking our learning about automation, getting the right advice, being aligned to the organisational goals and resourcing these automation programmes correctly is vital. But it’s about more than that. The parallels with Microsoft Access are undeniable to me – and in analysing the downfalls of the way so many adopted Access, we can see what we should (and really shouldn’t!) be doing now for RPA success.

Invest in the learning and people upfront

Just as Microsoft Access had too many people believing that they could develop software, robotic process automation isn’t just anyone’s game. Too many organisations are throwing juniors with a manual at the high-stakes task of building digital workforce capability. Many think that a short online training course equips them with all they need to know to get this right. Unfortunately, when people without the right skills launch into RPA, you end up with inefficient and ineffective automated processes that are badly designed and don’t deliver the intended business value. In some cases, these mistakes can even prevent an organisation from scaling their digital workforce – the very poisoned chalice scenario that we’re aiming to avoid.

While there’s no one part of building digital workforce capability that is especially difficult or complex, there are so many small pieces that are absolutely vital to get right. Because the thinking around RPA is still in its infancy, there’s a real shortage of people with enough experience. And while we could do with adopting more of the website or .net practice of downloading ready-made components that exist for RPA, as opposed to building it all from scratch, in a space that is still evolving, the right advice and focused expertise is still vital.

The business of automation has multiple bedfellows

With Microsoft Access, we saw operational people and department managers dive into utilising the tool to solve their specific problems, with little thought or reflection beyond an immediate fix. In some ways the approach here had merit, as automation isn’t (and shouldn’t be) strictly the domain of IT departments. Instead, it should start with, and be driven by, the business more broadly – those who are closest to, and best understand, the areas of need.

A mistake I’ve seen too many organisations make is treating the building of digital workforce capability strictly has a software delivery project, when it requires much bigger picture thinking. Any organisation looking at building a digital workforce should be considering HR, internal communications, risk, and change management to ensure that it will achieve what it needs to, while also taking an IT lens to the entire system to avoid a mess that needs cleaning up later. 

Combining the two approaches ensures that the organisation will not only have considered the functional requirements (i.e. the specific process that will be automated), but will also have taken into account non-functional requirements – the criteria that ensures your digital workforce will be able to perform those functions effectively. What security protocols need to be put in place? How have back-ups and maintenance or updates been considered? Have you considered availability? For example, does the robot still run if some aspect of the platform, or an adjacent system, falls over for whatever reason? IT can help the automation champions within a business cover more of these bases, and automation teams definitely need to work closely with them.

Design thinking and documentation

In the same way that people ploughed into Access to solve their most immediate business challenges, the temptation is to get stuck in here and just start building robots. I’ve talked already about a successful digital workforce requiring big picture thinking, but I’m going to make the point again; to truly deliver business value, it’s not a tactical thing to implement but instead a strategically considered programme that requires robust thinking.

Badly built processes that lack foresight require a lot of work to maintain. And, understandably, automated processes that aren’t documented well because they’ve come to fruition as people have felt (or fumbled) their way along are a nightmare. I know of several organisations that have had to retain specific people because they are the only ones who know how to use or support their current Access-derived systems.   

Designing something streamlined and simple from the beginning that performs as it needs to can feel like an idealistic luxury, not to mention the fact that it’s inherently more difficult. That’s the thinking that RPA benefits from though – and, when applied, it makes for a more robust and maintainable automation programme.

Understandably, there is a learning process involved in conducting a pilot, however there should ideally be a pathway that includes an increase in maturity to move an organisation from there to an effective automation programme. Start with mapping out processes – not only the first one being automated, but also other adjacent ones in the pipeline. This enables you to consider where you can create replicable components and make future development easier. The narrow and more short-sighted Access approach often meant that business challenges were being solved in isolation, which highlights the vital importance of picking the right problems to solve from the outset.

There’s so much more to say on the aspect of process selection that we will share another article soon solely exploring that aspect. However – suffice to say for now – those who are winning with automation are creating replicable components that can be adopted by later processes to make their subsequent development more viable. In this way, a successful digital workforce is able to achieve scale. It has the foundations to give it a better chance of keeping up with, or ahead of, the rate of change in business.

When you think you’re done, it’s not ‘set and forget’

Many Access systems were propped up in organisations to perform a specific function and then people washed their hands of them. They were just left to do their thing… until they couldn’t anymore. Too many organisations I have worked with found that once they upgraded Windows, their Access databases would no longer run, spurring a flurry of fixups marred by transitional down-time.

It wasn’t the best approach in the old days, but when it comes to building digital workforce capability, this ‘set and forget’ mentality simply won’t work. Short-term benefits should be used to set the stage for long-term success.

We know that the automation space and subsequently the ways we’re working are changing rapidly. So for a digital workforce to ultimately succeed and continue delivering business value, it needs to mature within the organisation so that its impact and reach can grow.

It largely comes down to experience

A recent survey by Pegasystems of 500 decision makers employing a range of RPA solutions (across various industries and across the globe) found that 52% said their robots were hard to deploy and 41% had difficulty maintaining them.

With many responding that their digital workforce required more time and resources to maintain than they expected, it’s clear that RPA is not an easy thing for just anyone to build and deploy. Mirroring many of my thoughts, the report asserts that successful automation programmes require “experienced automation experts, extensive investments in training, and ongoing collaboration between business and IT to be successful.”

Ultimately, the investment in experience is worth it. As with the Access days of old, we’re already seeing that digital workforce systems built badly are actually stopping organisations from scaling. They should be creating efficiencies, achieving positive return on investment through cost savings, and enabling increased agility. While in many cases they are, sadly some automation programmes are being switched off as they struggle to deliver business value.