Why is accurate sales forecasting such a challenge?

Why is accurate sales forecasting such a challenge?:from Building Scalable Businesses: The Inflexion-Point Blog by Bob Apollo

According to the latest research published by CSO Insights, less than half of all forecasted sales opportunities actually result in a sales win. Just under a third result in a competitive loss, and nearly a quarter result in the prospect deciding to do “nothing”. Just think about that for a moment. At a deal-by-deal level, the current state of sales forecasting in the average B2B environment is less reliable than simply tossing a coin.
Even at overall sales forecast level, when you might hope that some of the individual errors and inaccuracies would balance out in the overall revenue number, many organisations still struggle to come up with consistently accurate company-wide revenue forecasts. Now, I’m the first to acknowledge that successful selling is a blend of art and science, and that truly unpredictable changes in circumstance can sometimes trip you up, there’s clearly something going seriously wrong here.

Some companies are managing to do much better

In fact, it is possible to do better - and potentially a great deal better, as proven by recent research by the Aberdeen Group. They found that the vast majority (97%) of companies that implemented a series of best-in-class forecasting processes achieved their sales quotas - compared to just 55% amongst those that did not.
The performance difference is striking, but this is not just a matter of getting a number right. When forecasts are regularly missed, employees loose confidence, people lose their jobs, investors loose their money, and - ultimately - companies fail.

The barriers to effective sales forecasting

Aberdeen’s research offers some important pointers to the root causes of the forecast accuracy problem. It suggests that the most significant barriers to effective sales forecasting are:

1. Sales people not having sufficient knowledge of the details of specific deals, and/or (nearly as bad) failing to enter that information into the sales forecasting system
2. A lack of personal accountability on the part of individual sales people as to their responsibilities for accurate sales forecasting
3. Sales people displaying over-confident, conservative or sandbagging behaviours in their personal forecasting
4. Management failure to define or enforce strict stage definitions, milestones and data entry standards
5. A general inability to understand or calculate the realistic probabilities and closing dates for current deals

Although only the 4^th barrier is specifically called out as a management failure, in truth these are all failures of management. As the old saying goes, if you don’t inspect it, you can’t expect it. In the remainder of this article, I want to share a handful of pragmatic, time-tested initiatives that can enable managers to adopt a best-practice approach to sales forecasting.

The focus of this guide

I want to set the scene by explaining that these techniques are particularly effective in high-value, complex sales environments with lengthy buying cycles where a statistical approach to sales forecasting is unlikely to be effective. I’m talking about environments where - if it is to be accurate - the sales forecast has to be built bottom-up on a deal-by-deal basis.
If this matches your environment, this article should enable you to identify a handful of practical, immediately actionable ideas that will start to improve the accuracy of your very next sales forecast.  But even if your average deal values are low enough and your volume of sales transactions are high enough to allow you to take a statistical approach to revenue forecasting, I still think you’ll find some of the concepts valuable.

Getting the basic foundations in place

There are four critical foundations that you must establish before you can expect to achieve anything approaching best-in-class sales forecast accuracy:

1. A clearly defined, well-documented and consistently applied sales process with explicit milestones or gates between each of the stages
2. Clearly defined and consistently applied progressive opportunity qualification criteria that are regularly revisited throughout the sales process
3. A CRM system that is configured to capture, report and analyse the key facts and figures that affect sales forecast accuracy
4. A environment in which sales people see themselves as personally accountable for fully understanding their sales opportunities and for delivering accurate sales forecasts

There’s a common thread running through these four key foundations: clear definitions, consistently applied, supported by effective systems and personal accountability. In fact, without these in place, it’s hard to imagine how any organisation could generate consistently accurate sales forecasts.
Let’s explore each of these factors in a little more detail:

Clearly-defined sales process

Most sales organisations have implemented some form of sales process with named stages. But if this is to be effective, these stages must be clearly and unambiguously defined - and consistently applied across the whole sales organisation, with no scope or tolerance for creative interpretation.
The transitions between stages are equally important, and yet often poorly defined. It’s critically important that you establish observable, evidence based milestones - preferably grounded in buyer behaviour rather than sales activity - and insist that they are rigorously applied. No opportunity should be promoted by the sales person to the next stage in the process unless the milestone can be proved to have been achieved.
By the way, the first time organisations apply these rules, there’s often a dramatic shift in the weight of the pipeline with lots of opportunities being demoted to an earlier stage. Some drop out altogether. But it’s much better that this happens there and then, rather than clinging to the false hope that opportunities are far more advanced than they really are.

Progressive opportunity qualification

Qualification is not - and should never be - a one-time event. Opportunities must be progressively re-qualified throughout the sales process, and certainly whenever an opportunity is a candidate for promotion to the next stage in your defined sales process.
Your qualification criteria will undoubtedly vary depending from one offering and market to another. They will evolve from one stage to the next. But your qualification questions must always seek to answer the following questions:

• Is the opportunity real?
• Are they really likely to buy?
• Are we really likely to win?
• Is the deal really worth winning?
• Where’s the evidence?

Your CRM system

The best place to capture the information you need to assess each sales opportunity is your CRM system - and the quality of that information has a critical bearing on your ability to generate an accurate sales forecast. If you want good information, you need to take pains to ensure that your sales people see the system as inherently useful to them, and not just a convenient administration system for you.
If you want enthusiastic acceptance, and not merely mute and minimal compliance, your CRM system has got to pass your sales people’s “what’s in it for me” test. Avoid asking them for more information than you (or they) really need to make smart, well-informed decisions. Make sure you use the information to provide opportunity-level coaching that helps your sales people to evolve winning strategies for each account.

Environment of Accountability

It’s hard to generate consistently accurate forecasts without an environment in which every participant is held accountable for their contribution. Let’s start with the quality of information you expect the sales person to know about every opportunity. Be clear about which fields you expect to be completed in the CRM system, and make sure that the sales person has a personal commitment to capturing timely, accurate information.
Of course, forecasting isn’t just about having the facts at your fingertips - it’s also about making informed judgements with the data you have to hand. So accountability must also extend to the sales person’s responsibility for making intelligent and thoughtful judgements about the timing of each opportunity and the probability of it closing.

Building your forecast

Assuming you’ve established these foundations, the process of building your forecast should start with asking each of your sales people to present their forecast for the relevant reporting period. They should classify each opportunity they believe they have the realistic potential to close as either a “commit” or a “upside”.

• The “commit” deals are those where all the evidence (and they should expect to be asked to provide this) points to a bookable order being received within the reporting period without any need for unnatural acts
• The “upside” deals are those where the evidence points to a realistic chance of a bookable order being received within the reporting period if all goes to plan (and, yes, they should expect to be interrogated with regard to the plan)

By the way, I’ve also seen some sales organisations include a “long shot” category for fast tracking opportunities that with a following wind and a great deal of luck might just be closable. Of course, you’ll want to keep your eyes on these, but my strong recommendation is that the value of these deals must be excluded from any current forecast.
You’ll want them to provide the expected deal value (erring on the conservative side if there is a range), the probability of the opportunity closing in the period, the expected close date, and the current stage the opportunity has reached in the sales process.
The value of using probability percentages as a guide to forecasting is close to zero if all you’ve done is to take the CRM vendor’s out of the box stage probability percentages without any adjustment. If you use stage-related percentages, then you must base them on the actual percentage of opportunities that have been shown to close from that stage in your specific sales environment. If you don’t have the figures, then make it an urgent priority to calculate them. And if you allow your sales people to over-ride the percentages, you must define what you mean by them (for example, is it the percentage chance the deal will close in the current period, or ever?)
The sales stage is particularly important, because it ought to be (if you’ve applied consistent sales process definitions) an issue of fact, whereas the probability of that individual deal closing will always tend to be more of an issue of judgement. You’ll also want to ensure that the projected close date is backed by some clear evidence of buying intent or a compelling event and consistent with the rate of progress (velocity) usually observed for an opportunity at that stage in the sales process.
Make sure that “close date” means the same thing to everyone involved in the process. Does it mean the date on which the decision is made or does it take into account approval cycles, potential delays in the prospect’s order raising system, etc? Has it been confirmed in writing (or an exchange of email) with the executive sponsor?
We’ll talk later about the importance of measuring deal velocity, but for the moment it’s worth observing that one of the most common “unnatural acts” is a sales person’s misguided hope that they can close an opportunity far faster than other similar winning opportunities have taken from the given sales stage.
With each fresh iteration of the forecast, you’ll want to know what’s changed since their last submission. What deals have been closed? Which ones have moved forward? Which ones appear to be stuck or moving backwards? Have any of the values changed and if so, why? Have any risk factors changed, and how have they responded? Have their overall “commit” and “best case” projections changed as a consequence?
You now have the basic foundations for your sales forecast. But you’d be very unwise to simply roll up the forecast numbers and present that as your overall forecast. Assuming that you’ve carefully tested the issues of fact at a deal-by-deal level, you’ll want to apply intelligent judgement to the numbers in front of you.

The sales person’s past behaviour

Like it or not, and no matter how carefully you attempt to establish consistent expectations, you’ll know that different sales people have different approaches to forecasting. Some will be optimists. Some will be pessimists. Others may even have a nasty habit of sandbagging deals. You’ll get to know these different behaviours and - if the sales people fail to respond to your coaching - to adjust for them over time.
Another factor to take into account is the sales person’s typical closing behaviours. Some will have a habit of getting the deals in as early in the quarter as possible. Others will seem to always be leaving matters to the last minute. You’ll want to take these factors into account as well.

Sales velocity

One of the important ways in which you can validate your raw sales forecast is to compare the progress required for each deal with the typical time that it takes winning opportunities to be closed from that stage in the sales process. Sales velocity metrics are tremendously helpful in flushing out projected sales wins that would require “unnatural acts” if they were to be achieved.
Calculating the average time spent in each stage by winning deals can also be invaluable in identifying opportunities that have become “stuck in stage” and may require targeted action to get them moving again - or indicate that they should be removed from the forecast.

Conversion rates

You’ll also want to validate and if necessary adjust your overall forecast by taking into account the historic conversion rates from each stage in the sales pipeline. If the current forecast implies an unusually high or low success rate compared to past performance, and unless some key factor has changed, you should consider amending the forecast accordingly.

External factors

You’ll also want to take into account any changes in circumstance. These might be at the deal level (changes in the prospect organisation are a classic red flag) or in the market as a whole - perhaps changes in your competitive environment or a major external trend or event that is likely to affect your prospects’ buying decisions.

Focus on change

As you refine and adjust your forecast during the course of the reporting period, you should pay particular attention to what’s changed since the last review:

• Check to ensure that any changes (such as a stage advance) the sales person was predicting with regard to individual opportunities have in fact happened. If not, you need to investigate whether they have as firm a grip of the opportunity as they claim
• Pay attention to opportunities where their situation hasn’t changed. What is the sales person’s specific action plan to progress the opportunity? What are they waiting for the prospect to do, and what’s the basis for their confidence that this will happen?

Technology to the rescue?

Keeping track of what’s changed in your CRM system can be challenging - particularly since most CRM systems are poorly architected for this purpose. Many sales leaders are forced to resort to exporting the data into spreadsheets and then “staring and comparing” to identify changes and their root causes.
It’s a situation that gets progressively more difficult to manage as the size and complexity of your pipeline grows. Fortunately, technology solutions to the problem are available. I’ve had particularly good experiences with the offerings from Cloud9 Analytics (www.cloud9analytics.com) and Salesclic (www.salesclic.com). But technology is not a magic wand - you have to have laid a solid foundation first.

Action plan

So - if you want to make sure that you are in the best possible position to generate consistently accurate sales forecasts, you must first make sure that your organisation has the necessary systems and structures in place:

• A clearly defined sales process
• Consistent opportunity qualification criteria
• A properly configured CRM system
• An environment of accountability

Carefully examine your past performance. Make sure you understand your average sales velocity and conversion rates - and how they vary between different offerings, markets, sales people and sales channels.
Last, but by no means least, make it clear that you expect your sales people to be on top of the detail of their sales opportunities at all times, and that you expect them to base their judgements on clear evidence rather than hope or supposition.