Calling all telcos: it’s time to stop the software self-sabotage

“Debacle”. “Chaos”. “Meltdown”.

Dramatic words like these have become quite a fixture in our news feeds in recent years. The reason? More often than not, it’s a software issue. Just last week, the cybersecurity firm CrowdStrike revealed that it had cut its revenue and profit forecasts following a global outage in July that affected millions of Microsoft devices. The fallout reached scores of organizations and their customers, leaving CrowdStrike with a big financial and reputational recovery job on its hands.

Stories like these are more common of course because software is in just about everything. More businesses are run on it, especially since COVID changed default operating models and ushered in the rise of remote and hybrid working. Today, every business decision triggers a technology decision, the results of which will be “enjoyed” for years to come. And the risk is that a single flaw can quickly escalate into a huge issue. If something goes wrong somewhere, it could easily take your car off the road , freeze your thermostat , leave you stranded at the airport or your bank payments in limbo , or – worst of all – set your TikTok follower count to zero ??.

Joking aside, the repercussions can be dire. For consumers and service users, there can be negative effects on their privacy, personal data and even their safety. For corporates, there’s downtime, squandered transactions, and additional security risks, not to mention loss of credibility and trust. Even if the worst doesn’t happen, as those of us who were around saw with the Millennium Bug , the very act of preparing for it can run into hundreds of billions of dollars.

All of these problems are extremely expensive (not to mention time- and resource-heavy) in themselves. But even without them, running a software-based business burns a massive hole in most budgets. According to CloudEagle , a software-as-a-service (SaaS) platform, software is now the third largest expense for many companies, after employee and office costs.

Just consider the fact that developing a mobile application will cost anywhere from $15,000 to $80,000+ , while a “standard” enterprise-grade software solution (is there such a thing?) could be ten times more expensive again. Add in license fees and waste , per-request change fees and leaving fees – all of which are astronomical in their own right – along with the lifetime cost of maintenance, updates, refactoring and support, and it’s no wonder the global SaaS industry is currently worth more than $300 billion .

In telecoms, it’s a massive challenge. Even the best, most efficient operators still spend close to 3% of their total revenues on IT, which includes a very large chunk on software. In addition, the network division will be buying software all the time and eating up a double-digit percentage of revenues on capex projects.

Then we have the perception – which still lingers in some companies – that (non-network) technology is not core to the business, meaning many IT functions and critical “black box” solutions have traditionally been outsourced, with catastrophic results. Tied to external providers with no in-house talent, support, or direct access to their own systems, if there is an issue or outage, you cannot fix it yourself, you have to call the provider (good luck with that)… All of this means that, too often, operators are not in control of their own destiny when it comes to deploying and maintaining the technologies that their businesses depend on.

Last but not least, as many experts have alluded to, the real problem with software development is not writing code, it’s managing complexity. And there is plenty of that in the telecoms industry. For a start, it is mired in consolidation, which as every telco CTO knows, is an absolute nightmare. Rolling out software across one set of crazy company systems is tricky enough. Mash two (or more) together and you are left with the IT equivalent of a Gordian knot: complex systems, cumbersome development processes, and painfully long testing cycles.

The result is ridiculously extended timeframes. Typically, it takes somewhere between 6-12 months to introduce something new to customers, and some companies need weeks or even months to do something as “simple” as a price change. Telco IT teams are the only people who get away with having a three-year functional plan, rather than a KPI-driven one, because everything takes so long. Imagine your chief of sales promising you 1000 new customers in five years’ time! So there are hundreds, if not thousands, of telco technology plans out there, not delivering any benefits to the business.

It means the business can only do a fraction of what it really wants or needs to, because the development roadmap and budget are always full. It also means when solutions are finally deployed, they are either out of date, no longer required, or don’t meet actual requirements. This is aggravated further by the business not understanding standard software capabilities, which drives costly tweaks and customization. To top it all off, requirements are routinely passed from pillar to post between business owners, project managers, the technology team, and suppliers, becoming more complex or diluted along the way.

Add in adoption rates, and how well you’ve managed the implementation (which could be a whole other article in its own right), and there are so many pain points it’s easy to see why companies in telco and beyond are awash with shadow IT : unauthorized solutions that have not been approved by their technology departments.

The result is a chaotic, costly, security-compromised mess. Really, it amounts to self-sabotage and nobody has genuinely managed to change things, despite some operators attempting to position themselves as technology companies.

In my view, it’s time for telcos to stop sabotaging themselves in this way, and get a handle on cost, control, and complexity instead. Ultimately, operators need to find an answer to the eternal software conundrum: how to create code that works reliably and can be updated and maintained by somebody other than the original authors. This needs affordable, flexible, secure solutions that:

• Are designed for telcos, by telco experts
• Let operators build up their own software capabilities over time

And, perhaps most importantly

• Enable them to deliver new products and services to customers, easily, and at speed

It’s been thirteen years since Marc Andreessen said software was “eating the world”. It shouldn’t be eating up telcos: it should be a tool for them to make the leaps and bounds we know are possible.

Photo credit: amgun, iStock