The Real Reason Leadership Doesn't Trust the Salesforce Forecast

For decades, organisations have invested heavily in improving the accuracy of their sales forecasts. Entire software categories have emerged around pipeline management, predictive analytics and revenue intelligence, each promising greater visibility into future performance. Salesforce alone has evolved from a customer relationship management platform into an increasingly sophisticated decision-support system, capable of analysing historical trends, identifying at-risk opportunities and even generating AI-assisted forecasts. On paper, there has never been a better time to produce reliable sales predictions. Yet confidence in those predictions appears to be moving in the opposite direction. Recent industry research suggests that only around one in three revenue leaders genuinely trust the forecasts presented to them. Other studies paint an equally concerning picture, showing that fewer than a quarter of organisations consistently achieve what would be considered an accurate forecast. Those figures should give every executive pause. Forecasting is not simply another management report. It influences hiring decisions, investment priorities, production planning, cash-flow management and conversations with investors. When leadership begins to question the credibility of the forecast, uncertainty spreads far beyond the sales department. The natural response has been to look towards technology for answers. If forecasts are unreliable, then perhaps the CRM requires refining. Perhaps opportunity stages need tightening, salespeople need to update records more consistently, or artificial intelligence can identify patterns that humans continue to miss. The market has responded enthusiastically to these assumptions. Organisations are encouraged to improve pipeline hygiene, strengthen CRM discipline, implement revenue intelligence platforms and invest in increasingly sophisticated forecasting tools. None of these recommendations are unreasonable. Clean data is preferable to poor data, consistent processes usually outperform inconsistent ones, and modern analytics can reveal trends that would otherwise stay hidden. The difficulty is that these solutions all begin from the same premise — that inaccurate forecasting is fundamentally a data problem. I would argue that this is where the conversation drifts away from reality. Poor forecasts are rarely created because Salesforce lacks information. They emerge because the information entering Salesforce has already been shaped by human judgement, organisational incentive and behavioural adaptation. By the time a forecast reaches an executive dashboard, it represents the accumulated outcome of hundreds, sometimes thousands, of individual decisions. Every opportunity has been assessed by a salesperson, interpreted by a manager, reviewed by a regional leader and, in many organisations, adjusted again before it reaches the boardroom. At every stage of that journey, people decide what to report, when to report it and, most importantly, how much of the truth feels safe to reveal. Seen from that angle, forecasting becomes something very different. Rather than a reporting exercise supported by technology, it becomes a reflection of organisational culture. The numbers inside Salesforce tell us as much about trust, incentive and leadership as they do about customers or revenue. That distinction matters because it changes where leaders should direct their attention. If forecasting is primarily a reporting problem, investment should focus on better reporting. If it is the visible outcome of behavioural patterns established across the organisation, dashboards alone can never solve it — they simply reflect decisions that have already been made elsewhere. Part of the reason this misunderstanding has persisted is that the language surrounding forecasting has become more technical than human. Browse almost any article on the subject and the same phrases appear: pipeline hygiene, CRM discipline, forecast categories, stage definitions, revenue intelligence. These terms sound reassuringly objective, suggesting forecasting is largely a question of process and system design. Take *pipeline hygiene*. It implies that poor forecasts result from untidy data that simply needs cleaning — opportunities updated more regularly, duplicate records removed, stale pipeline eliminated. Few would disagree with those recommendations. But the phrase also carries an assumption that rarely gets challenged: that data has somehow become dirty through neglect. In reality, data doesn't deteriorate on its own. It deteriorates because people make conscious decisions about what they choose to record. That difference changes the nature of the problem entirely. If a salesperson repeatedly delays updating opportunities, the obvious explanation is a lack of discipline. But another explanation exists: perhaps experience has taught them that updating opportunities too early exposes them to criticism if circumstances change. When probabilities appear consistently inflated, it's tempting to conclude salespeople are simply unrealistic. An equally plausible explanation is that optimism has become culturally rewarded while caution has become professionally risky. Most organisations ask why salespeople aren't updating Salesforce correctly. Far fewer ask why capable, experienced professionals might have perfectly rational reasons for behaving exactly as they do. That's the distinction that shifts responsibility away from software and towards organisational behaviour. Salespeople interpret buying signals, assess customer intent, estimate commercial risk and decide whether an opportunity has genuinely progressed. Managers review those assessments through the lens of their own experience. Regional leaders add another layer of interpretation, balancing commercial optimism against board expectations. By the time the executive team reviews the final forecast, the numbers represent a series of interconnected human opinions rather than a purely objective read of future revenue. Technology can organise those opinions remarkably well. It cannot make them objective. Optimism bias, for example, is almost inevitable in high-performing sales teams. The same qualities that make salespeople effective — the belief that obstacles can be overcome, relationships strengthened, negotiations rescued — also encourage them to see positive outcomes more readily than negative ones. A customer expressing interest can easily become interpreted as commitment. Within sales circles, this tendency has earned its own label: happy ears. Salespeople hear what they hope to hear rather than what has actually been said. This rarely stems from dishonesty. It is optimism expressed through commercial judgement. But when hundreds of opportunities are shaped by that same optimism, the cumulative effect can distort the forecast without anyone deliberately manipulating a single record. Process may determine how opportunities move through Salesforce. Human psychology determines how people decide to move them. That distinction becomes even clearer with one of the oldest behaviours in sales management: sandbagging. Almost every experienced sales leader knows the term — salespeople deliberately understating likely revenue, or delaying committing opportunities until they feel almost certain a deal will close. Conventional wisdom treats this as poor forecasting discipline. I think that explanation stops far too early. Sandbagging is rarely a technology problem. It is a behavioural adaptation to organisational incentives. Imagine two organisations facing the same commercial reality. In both, a salesperson includes a significant opportunity in the quarterly forecast. The customer's procurement process slips, pushing the deal into the following quarter. The revenue outcome is identical in both. The leadership response is not. In the first organisation, the salesperson is challenged publicly during the forecast review, their judgement questioned, their credibility quietly diminished. Although the delay sat largely outside their control, the experience leaves a lasting impression: forecasting confidently carries personal risk. In the second, the manager approaches it differently — exploring why the deal moved, what early signs might have been missed, how future qualification could improve. The missed forecast is still taken seriously, but it becomes a coaching opportunity rather than a verdict. Both organisations experienced exactly the same commercial outcome. Only one strengthened trust. Over time, each business builds a different relationship with uncertainty. In the first, salespeople learn that committing revenue too early can damage their reputation, so the rational response is caution — delaying commitments until success feels guaranteed. Forecast accuracy deteriorates, not because people are incapable of forecasting, but because the organisation has taught them that caution is safer than honesty. In the second, uncertainty is treated as an inevitable part of selling rather than evidence of poor judgement, so risks surface earlier, coaching happens sooner, and leadership gets a more realistic view of the commercial landscape. Ironically, by becoming more accepting of uncertainty, that organisation produces the more reliable forecast. Sandbagging isn't a forecasting failure. It's evidence that people have adapted perfectly to the environment leadership created. If protecting personal credibility becomes more important than flagging risk early, that is exactly the behaviour the organisation should expect to see. This is one of the reasons organisations so often describe forecasting as being "more art than science." It's usually meant as a harmless observation, but it reveals something more significant than it intends to. Forecasting becomes an art precisely because trust has been replaced by interpretation — every layer of management applying its own judgement to compensate for perceived weaknesses elsewhere in the organisation. The result isn't a more accurate forecast. It's simply a more complicated one. This also challenges an assumption that gets little scrutiny: that distrust flows only one way, from leadership down to the sales team. In reality, sales teams frequently stop trusting leadership long before leadership stops trusting them. Consider what happens when executive teams inflate targets to satisfy the board, or managers quietly adjust the numbers they receive because they assume every rep is being optimistic. At each level, people compensate for what they believe everyone else is doing. Salespeople become conservative because they expect management to raise the number anyway. Managers discount opportunities because they assume optimism is already baked in. Senior leaders apply their own adjustments before presenting externally because they no longer believe the internal figures. Forecasting stops being an exercise in understanding reality and becomes an exercise in managing expectations. Each individual believes they're correcting someone else's bias. Collectively, the organisation drifts further from an honest read of the market. The final forecast becomes a negotiated position rather than a genuine prediction. At this point, it's worth asking an uncomfortable question. If leadership doesn't trust the information inside Salesforce, is Salesforce actually the problem? In my experience, the answer is usually no. Salesforce has become remarkably good at recording information — customer interactions, opportunity movement, activity levels, extraordinary visibility into commercial performance. But the platform can only record what people choose to give it. If opportunities are updated late, Salesforce reflects late updates. If stage definitions are interpreted differently by different teams, it records inconsistent progression. If coaching happens in notebooks, spreadsheets or informal conversations rather than inside the platform, Salesforce has no way of capturing that knowledge. It simply mirrors the behaviour taking place around it. A CRM system cannot establish psychological safety. It cannot remove fear from a forecast review, or persuade a leader to respond more constructively when a deal slips. Those are leadership responsibilities. Salesforce merely makes the consequences of them visible. When leaders say they can't trust their CRM data, what they're often looking at isn't a technology failure — it's an honest reflection of the culture that produced it. Replacing the mirror rarely changes the reflection. This is also why so many organisations invest heavily in forecasting technology without a corresponding rise in confidence. Artificial intelligence is the clearest example. Machine learning can identify patterns and surface risks that might otherwise stay hidden — but it cannot remove the behavioural distortions already embedded in the data it learns from. If history contains inflated probabilities, delayed updates and widespread sandbagging, the model learns from exactly that. It may describe the distortion with remarkable sophistication. It cannot turn it into objective truth. Better analysis of poor inputs still produces unreliable outputs. That isn't an argument against AI — it's an argument against expecting technology to compensate indefinitely for organisational behaviour. Leaders often hope a better algorithm will solve a problem that is fundamentally human. It rarely does. Organisations that forecast well rarely start by talking about forecasting at all. They build consistent qualification criteria, not for the sake of compliance, but because shared definitions reduce ambiguity. They train managers to coach rather than simply inspect pipeline reports. Above all, they separate uncertainty from incompetence — recognising that selling has always involved things outside a salesperson's control, and that the objective isn't to eliminate uncertainty but to surface it as early as possible. As trust increases, people become more willing to share uncertainty honestly. As uncertainty becomes visible earlier, coaching improves. Better forecasts become a consequence of that, rather than the primary target. Forecasting accuracy is not created in the quarterly forecast meeting. By the time executives gather around the boardroom table, the quality of the forecast has already been determined by hundreds of conversations that took place over the preceding weeks — in one-to-one coaching sessions, pipeline reviews, and informal conversations between managers and their teams. It was shaped every time someone decided whether it felt safe to admit a deal was at risk. The executive dashboard simply reveals the outcome. That's why so many organisations remain disappointed with their forecasting initiatives. They keep searching for better dashboards when the real opportunity lies in better conversations. They invest in increasingly sophisticated technology while leaving the behavioural system feeding that technology largely unchanged. Salesforce has never been more capable of producing accurate forecasts, yet confidence in those forecasts continues to decline. The limiting factor is no longer the software. It is the quality of the human relationships surrounding it. The next time a sales forecast appears on the boardroom screen and someone asks whether the numbers can be trusted, it may be worth resisting the instinct to examine the dashboard more closely, and asking a different question instead. What have we taught our people is safe to tell us? That question reaches far beyond Salesforce, into organisational culture, management behaviour and leadership itself. Every company has a forecasting system, but very few recognise where it truly exists. It does not begin with opportunity stages or probability percentages. It begins with the environment leaders create every day — through the questions they ask, the behaviours they reward, and how they respond when reality proves uncomfortable. Long before a revenue figure appears on an executive dashboard, someone has already decided how much of the truth feels safe to share. That decision — not the CRM configuration, not the quality of the reports, not the sophistication of the AI model — is what ultimately determines whether leadership trusts the forecast. Perhaps the future of forecasting will not belong to the organisations with the smartest technology after all. Perhaps it will belong to those that build cultures where telling the truth is no longer the riskiest thing a salesperson can do.