Vention's interactive State of AI report is waiting for you! Check it out for the lowdown on AI funding and which sectors and companies are leading the pack. Get it now.
Visit EN website
dark modedark modedark mode
light modelight modelight mode
Light mode
Last updated: May 24, 2023

Lessons learned from a seasoned CTO

Author: Ellie Jones
Last updated: May 24, 2023
What's inside

Even the most successful CTOs have made their fair share of mistakes along the way. Skin-thickening, character-building . . . call it what you will, but in the end it’s all about learning.

And it’s not just about learning from your own mistakes. One of the great things about having a friendly professional community is that we can share our lessons learned.

At the 2023 CTO Craft Con, our CTO of Digital Solutions, Glyn Roberts, shed light on common mistakes tech leaders make — and how to learn from them going forward.


Lesson #1: Slow and steady wins the race

Solutions that are built quickly can bring in quick cash, so it can be tempting to prioritise speed, even when there is a likelihood of incurring technical debt. But leaders need to keep their eye on the prize: a solution’s overall stability and capacity to scale.

Resist the temptation to overhaul a solution or system completely unless it’s absolutely necessary. “A total rebuild happens a lot in product companies that try to act like an agency to meet certain clients’ requirements rather than focusing on their global product offering,” says Glyn. “Though this approach is sometimes necessary, relying on it for an extended period will take its toll.”

Big-bang rebuilds are risky in other ways, too. As Glyn notes, when you try to push through a total rebuild — especially one that isn’t vital — it becomes increasingly likely that something will be missed and mistakes will be made, even with the best planning and top-notch developers.

Moreover, even if you’re trying to complete a full revamp quickly, you can lose valuable time. “The timelines to achieve an overhaul can be long, so the delay to seeing value is too long,” Glyn says.

Tackle fixes to a solution in a smaller, more iterative way, if you can. By assessing and stabilising your existing solution, you’ll wind up with one that is more manageable, scalable, and reliable, with less downtime, minimal risk, and faster investment to value.

Lesson #2: Mix ‘sexy’ with sensible

Don’t fall prey to the shiny new thing without making absolutely sure it will work in your favour.

“Given the power to make decisions without effective oversight, inexperienced engineers will invariably opt to work with what’s ‘sexy,’” says Glyn. That’s understandable, given that trending technologies aren’t just exciting — they’re CV gold.

But when these solutions are unsuited to product requirements, you’re left with not only a technical problem but a management problem: In this scenario, you risk having engineers pack up with their new skills to find a new job. Then your company is left not only with a troublesome product, but no one on hand to fix it either.

That’s not to say that you shouldn’t embrace new technologies, but you should do so in a managed way, building on existing solutions and implementing analytics for longevity. “Boring is less sexy,” Glyn acknowledges, “but much easier to maintain and grow as needed.”

Lesson #3: Embrace ideas that could improve operations

The more complex the product you’re building, the more time will be required, and the higher overhead will be. Overall, this will result in more man-hours, higher risk of human error, and smaller profits. Moreover, if a solution is complex but doesn’t have broad-reaching impact, then its value is limited.

The answer is simple: Implement a technical solution that improves operational activities, reducing the number of people needed and minimising the risk of mistakes.  Though the initial outlay can be significant, an effective digital solution will be more cost-effective over the long haul. “The earlier you implement your solution,” Glyn says, “the faster value is delivered to the business.”

One of the core functions of a CTO (or any leader) is to make tough calls and think strategically, both in the moment, and longer-term. Without a strategy that guides every move — big picture thinking — things can fall apart fast. A single-minded, short-term plan aimed at quick wins is tantamount to no strategy at all.

Keep reading:

Best practices for chatbot development
Customers are increasingly expecting bots to answer questions and grease the wheels for purchases. Training a bot well and integrating it with solutions like Salesforce can transform your business.
Mar 30, 2023
How to handle challenges in artificial intelligence
Rapid developments in AI are ushering in an IT revolution. A keen awareness of the many ways AI can go off the rails is paramount.
Feb 6, 2023
Python or Julia for data science: Which one works best?
Let’s explore Julia performance in comparison to Python’s to learn which one, Julia vs Python, is a better option for data science?
Nov 10, 2020