Customer Relationship Management (CRM) was supposed to bring order to chaos. But instead, it’s become one of the biggest silent killers of early-stage companies, argues Elliott Limb at Cubed

The average lifespan of a Chief Revenue Officer is less than two years. That’s less than the life of a hamster. The startups they preside over don’t fare much better: 90% of startups will fail within three years and only 2% of scaleups make it to Series D. Overall, three-quarters of venture investments bite the dust.
If you say these numbers scream that something’s fundamentally broken within startups, you’d be right. It is. The Customer Relationship Management (CRM) model, the engine at the heart of a startup, is consistently failing its users. It’s not driving revenue - it’s driving failure. And it’s high time we had something new.
Dysfunction by design and the culture of bad data
When Salesforce first came to the UK in the noughties, its cloud-based CRM was a breath of fresh air. Compared to the clunky installed software they were previously using, it offered sales teams significant advantages and was easy to scale.
But over time that story changed. To add much-needed functionality, businesses found themselves seeking out additional software to fulfil crucial tasks, such as improved forecasting or call recording. The result? An unwieldy, Frankensteinian monster that’s hard to control and hungry for company profits.
Compounding the problem is the culture of bad data that’s sprung up within sales teams. Sales representatives - under pressure to drag deals forward and forced to fill in endless data fields - log optimistic guesses to appease their managers. Revenue leaders, straining to meet ambitious sales targets and roadmaps, pass this puffed up data along the chain. Senior leadership believes this forecast information - and hands it over to investors. These inflated pipelines and forecasts mislead investors and derail strategic decisions. Our analysis of 375 qualifying questions, across 100 scaling VC-backed businesses, reveal a culture of guesswork dressed as strategy. No wonder the startup and scaleup failure rate is so spectacular.
Bad learning
Another huge issue for startups lies in the way they train their representatives. Traditional sales enablement doesn’t work: classroom style training is an ineffective teaching tool, even if you do manage to get all your reps to attend. Startups still rely on one-off training sessions, battle cards, and a ‘one-size-fits-all’ approach - expecting this to somehow appeal to the most individually-minded reps in their teams. They have the data - even if it is inaccurate - on individual rep activity, but this isn’t abstracted or used to inform individual, tailored training. And so reps aren’t empowered to reach their full potential and the business continues to underperform.
So much more could be done, in terms of tailored feedback, coaching and guidance across handling objections, pricing, contractual negotiations and other parts of the sales process if there was better data about deals, and better information about competitors and competitive products or services available during sales conversations, not in one-off, isolated training sessions.
AI can rebuild from the ground up
If the CRM is broken, how do we fix it? Slapping new tools on top of bad systems won’t work - even if they do claim to use AI. The solution isn’t pretty but it’s the only way you’ll break the cycle of failure. You’ve got to gut your CRM and replace it with something entirely new.
A new revenue engine, with artificial intelligence embedded in its core, is the solution to the bad data and bad learning that’s been plaguing startups for years. Imagine never having to rely on manual data entry, because your AI-enabled engine can capture real-time signals from meetings and emails, sales conversations, and Companies House data, among other sources. Not only does this make reps’ lives easier - who got into sales because they want to fill in 50 ‘mandatory’ fields for each order form? - it puts an end to the culture of guesswork that’s tanking startups’ success. No more hopeful estimates that do more harm than good: startups can get accurate, real-time data that helps them identify gaps and make better strategic decisions. Furthermore, forecasting is transformed if data is captured automatically, via configurable AI which ‘learns’ from every prospective customer interaction, every instance of deal slippage, and every time a forecasted deal has fallen through.
Sales coaching and enablement can also be completely reimagined. AI-enabled tools can prompt sales reps to ask the right questions at the right time in live sales conversations, and help them respond to nuanced objections in the moment, rather than reading off a stale sales script. Such support not only helps close deals but gives reps continuous on-the-job upskilling and training, without having to have a manager stand by their side. Live, contextual coaching, embedded in the flow of work, is in a completely different league to tired classroom training.
Startups have become trapped in a cycle of failure. There’s not any one culprit in this destruction of value: reps, managers and leaders are all trying to do their best, but they’re wrestling with an engine that’s geared to work against them. That engine needs to go. It’s only by replacing the current CRM model, and switching to a totally new operating system, that they’ll generate the figures they so desperately want to see.
Elliott Limb is CEO of fintech growth partner Cubed
Main image courtesy of iStockPhoto.com and GamePH

© 2025, Lyonsdown Limited. Business Reporter® is a registered trademark of Lyonsdown Ltd. VAT registration number: 830519543