3aIT Blog


As a provider of various IT services, we often get questioned as to why we're recommending a new server or to ditch an old application critical to your business in favour of a new system that makes use of the latest technologies to improve efficiency and security. We completely understand the scepticism. Of course we're going to say this - replacing old systems accounts for a fair chunk of our income! However, we try to be as objective as possible when giving advice along these lines. We will never try and sell a new system to someone unless we genuinely think it will improve things.

The drip-drip of small improvements

The root cause of this is mostly due to the slow evolution of PCs. One of the biggest upsides of PCs (especially those running Windows) is also one of the biggest downsides. The thing we use to do large chunks of our jobs doesn't seem to  have changed in about 20 years. You press a button on the front of a box that probably used to be a bit bigger and noiser, but is basically the same, and that sends a picture to a monitor that used to be a lot bulkier and squarer, but again is basically the same. Once it's loaded, you probably use Word, email and websites in basically the same way you did when you first used a PC.

So because you're sitting in front of what seems to be the same thing you were sitting in front of 20 years ago, the expectation is often that the things that worked 20 years ago should still work today. However, both the hardware and software that powers that box of tricks has evolved massively over that time. It's just happened slowly, so you rarely notice.

Therefore, the issue we face as suppliers of both hardware and software is that there's no cut-off point that we can concentrate all our efforts on the new. You wouldn't complain to the manufacturer of your Smart TV that it can't play VHS tapes. Because we've just bought a new device that is distinctly different to the thing it replaced, we accept that things have moved on and adapt accordingly.

It works. Why change it?

With PCs however, people expect to be able to plug in a printer they bought in 2001 and to be able to run a rickety old Access based database that was either bought off-the-shelf or written for them at around the same time on a PC that was bought yesterday.

And a lot of the time, it does "sort of" work. Because things have changed slowly, various processes are introduced to workaround issues with ancient systems bit by bit. Microsoft goes to great lengths to try and ensure backwards compatibility in Windows. While this is often helpful, it does mean there's no definitive point that old hardware and software has to be upgraded, so the tendancy is to try and stick with it for as long as possible.

While a system may well have been the best available solution at time, 10 years down the line both your business and technology has moved on. As this has all happened slowly, one can find themselves wasting half their time fighting an old bit of software without even realising it. If you were to have suggested the system they're now battling with to them when they bought it, they would have laughed you out of the door. However, because the staff all know how it works, there's often a lot of resistance to the short-term disruption of upgrading to something better.

The consequence of this is that the decision to do something about it is left until the last possible moment, and at that point, an alternative needs to be found *now*, greatly limiting the opportunities to engineer exactly the right tool for the job. Instead, people will opt for something that does 80% of what they need, put processes in place that work round the other 20%, then call the job done and learn to live with deficiencies. This builds in wasted man-hours from day one, and then the cycle starts again until this new system falls apart at a critical moment years down the line.

If a system is upgraded in a forest and no-one hears it, did it actually happen?

The other problem we face here is that a lot of the improvements that get made in our line of work are basically invisible to the end users. This is especially the case when it comes to security concerns. We can certainly understand the reluctance to replace an old Windows 2008 server with a new one when the users will see little to no difference. The shared files will look the same. They won't load any quicker. And yet doing this is something that should be in the planning now for anyone still using it. Software cannot be supported forever, and when support is withdrawn, your business will instantly become vulnerable to a whole host of issues that could bring your business down for days.

This is why we weren't surprised by this BBC report the other day that 1 in 5 firms have been hit by cyber attacks. We suspect this may be even higher for firms below the 100 members of staff level that was measured here. Time and time again we see security ignored until it's too late.

Trusted advice on evolution

So what's the answer? Firstly, find an IT company that you trust to provide good advice. Rather than being wowed by a list of clients that they've dealt with at some point in their history, ask what percentage of them have been with them for years. This is a good indication of a relationship that works both ways rather than an IT company that flits from one customer to the next selling whatever the new shiny thing is that they're currently selling, and then moving on.

Once you've found a provider that you trust, it will be much easier to act on the advice they provide without worrying they're taking you for a ride.

As far as hardware is concerned, replacing this every few years is something that just needs to be costed in. A lot of the time, it is often a blessing for the staff when their PC breaks. We frequently see ancient machines still in service that take half an hour to boot and 10 minutes to load a web page. These staff members are losing at least an hour a day just waiting for their machine to respond. And yet because it "sort of" works, they're stuck with it until it finally splutters and dies.

As a rule, PCs that are used every day should probably be replaced every 5 years or so. The lifespan for other hardware varies depending on what it is, what it's used for and how often it's used. Again, knowing when to replace this goes back to dealing with a company you can trust to let you know when it's time to replace something.

On the software side, our advice usually to move to a web-based system. Tying your business to a specific bit of code (eg a Payroll system or CRM) that has to be run locally on each PC that's using it is just numbering the days until the hardware that can run that bit of code no longer exists, and you have no choice but to ditch it.

The beauty of a web-based system is that it can evolve in parallel with the technology. Rather than looking for a system that does everything you can possibly think of right now, it can be a lot more productive to start with a core system that covers all the essential bases, and then slowly add to that over time. This way, you don't need to worry about what you might need five years down the line on day one. Four years down the line, you can work with your IT provider on that feature, and it'll naturally use all the latest technologies that didn't even exist when the system was originally built.

This applies to websites too. So often we see websites that were built to spec a few years ago, and then not touched since. Not only is this a huge security risk (these are often running on ancient versions of Wordpress), it also means they quickly look out of date compared to the competition. Again, a trusted company can help you keep these up-to-date and suggest ways you can evolve your website rather than rebuild it from scratch every few years.

The short version...

So, in short(!), you will find IT much less frustrating if don't resist moving with the times. Just because something is "basically" working now, it doesn't mean you should stick with it until it's completely broken. The price of no progress creeps up on you slowly until you realise too late that you've made a mistake not addressing it sooner.