Lean B2B sales process (lean sales): how to remove the steps that do not lead to a contract

In my analyses with sales and marketing teams, during the support I provide inside companies, I hear one question return:

“How do we make our sales process lean, by removing the steps that do not lead to a contract?”

It is a theme that, periodically, comes back in vogue in the fourth quarter. I bring it here now for a reason of calendar:

• we are in the fourth quarter, the one in which the objective of the year must be reached;
• a useless step weighs more, because the time it consumes cannot be recovered in the following quarter.

This article is addressed to these decision-making roles:

• Sales: Sales Manager, Area Manager, Sales Director;
• Marketing: Marketing Manager, Marketing Director;
• Smarketing: sales and marketing as one team on the same revenue objective.

The short answer

To make a B2B sales process lean (that is, free from waste) three ingredients are needed, and they do not weigh the same:

1. Artificial intelligence: about 20%. It is the tool that makes it possible to apply the method at a human speed.
2. Data: about 30%. Theoretical data (from the market) is distinguished from pragmatic data (internal to the company).
3. The internal knowledge of the company: about 50%. It is the heritage of values, processes and strategies that nobody else owns.

It is the proportion I observe in my work with companies: a rule of thumb, not a statistic. The rest of the article demonstrates it.

Who decides what on this theme

Role	What it decides on this theme
Sales Director	Which targets to hit and with which process
Sales Manager	Which steps the salespeople really perform
Area Manager	Where the process slows down in the territory
Marketing Manager and Marketing Director	Which content accompanies each step

The problem: which step does not lead to a contract?

The sales process (the sequence of steps between the first contact and the signature) becomes longer by accumulation. These are the typical culprits:

• a form introduced after an unlucky episode;
• an approval requested “to be safe”;
• an alignment meeting that nobody has the courage to cancel.

The Lean Sales approach (the application of lean thinking to the sales process) offers a criterion: a step is worth the value it has for the buyer.

But among all the steps that do not lead to a contract, one is fundamental, and it must be highlighted before the others:

The choice of the profile of the target, that is of the targets to identify and hit, made on non-analytical criteria. It happens when the internal signals of the company are not captured in depth, because the strategic approaches and the operational tools for a deep analysis of the target, in relation to the objective to be reached, are missing.

A badly chosen target makes every following step useless, however well executed. Then comes the rest, the process weighed down by the usual things: forms, approvals, meetings.

Four questions to recognise a step to remove

Control question	If the answer is yes	Decision
Does the step give value to the buyer?	The customer perceives and appreciates it	Keep it
If removed, does the probability of signature decrease?	The step affects the result	Keep it
Does it exist only because of a past episode?	No data justifies it today	Put it to the test
Does the customer endure it without gaining anything?	It is waiting, not value	Remove it

How do we decide, then? With data, and here begins the part that almost everybody gets wrong.

Two kinds of data: theoretical and pragmatic

To identify the strategic objectives I always distinguish two families of data.

Theoretical data and pragmatic data compared

	Theoretical data	Pragmatic data
What they are	Market data, almost always from third parties	Data internal to the company, the first-party data
What they measure	How specific we are compared with the market and where we are positioned	How we really work, step by step
What they are for	Benchmarking (the comparison with market values)	Deciding which target to hit and which content to add, remove or change
When to use them	To orient ourselves, and when internal data is missing	Always, 24 hours a day, 7 days a week, 365 days a year, as the main raw material

An example from the car trade. Market data says that some brands sell faster because they adopt affordable pricing models. If we sell cars, the comparison has two outcomes:

• our times are longer than the market’s: we are slower than competitors;
• our times are shorter: we do better.

This is theoretical data: it says where we stand compared with others, but it does not say why we are slower than our competitors (and being slower is not necessarily bad).

An example. If our competitors work with standard technologies and we work with cutting-edge technologies, it is inevitable, and also correct, that our target takes longer. Choosing us means weighing an added value that others do not offer, and it is our winning factor. Slow, therefore, is not always bad. But this is told to us by pragmatic data.

The why is in the pragmatic data, which say:

• at which point of our negotiation the customer stops;
• after how many days;
• with which kind of offer.

Here is your heritage. Theoretical data is taken from outside only when internal data is not sufficient, because it is blurred, incomplete, or because we did not keep it. A company that did not preserve its commercial history has lost critical mass, and critical mass cannot be bought.

The ten Lean Analytics rules to decide what to remove

Lean Analytics is the book by Alistair Croll and Benjamin Yoskovitz (O’Reilly, 2013) from which the measurement method described in the Lean Analytics glossary entry takes its name. Born for startups, it works for any process that must be made simpler. I derive ten rules and apply them to sales. The examples with numbers are invented.

The ten Lean Analytics rules applied to sales

No.	Rule	Control question	Example in sales
1	Comparative	High or low compared with what?	18% of quotations becoming contracts, compared with the previous quarter
2	Understandable	Does the team remember and discuss it?	"Days from first contact to quotation"
3	A ratio or a rate	Does it contain a before and an after?	Quotations closed out of one hundred sent, not quotations sent
4	Changes behaviour	If it rose or fell, would I change something?	If the answer is no, it leaves the dashboard
5	Actionable, not vanity	Does it lead me to do something?	Calls that produce an appointment, not calls made
6	Anticipates or certifies	Do I still have time to act?	Qualified contacts (anticipates), signed contracts (certifies)
7	Qualitative and quantitative	Did I listen before counting?	Ten lost customers interviewed, then a count of the reasons
8	Explores or reports	Am I searching or controlling?	Removing a step is an experiment, not a report
9	Separates causes from symptoms	Am I treating the symptom?	Quotations without reply: cause of the problem or symptom of a customer not convinced
10	One per stage, not too many	Is it indispensable now?	This quarter the time to quotation, the next the closing share

1. A good metric is comparative. A metric that cannot be compared is a vanity metric (a number that feels good and does not say what to do). The comparison with the market is the task of theoretical data.
2. It is understandable. If the team cannot remember it and discuss it in a meeting, it will not change any behaviour.
3. It is a ratio or a rate. A ratio contains a before and an after, so it is comparative by nature.
4. It changes behaviour. The test is a question: if this number increased or decreased, would I change anything in what I do? If the answer is no, remove the metric from the dashboard.
5. It is actionable. An actionable metric leads you to do something. Calls made is vanity; calls that produce an appointment is actionable.
6. It anticipates the future or certifies the past. Lagging metrics tell the past: contracts signed in the quarter. Leading metrics allow action while there is still time: new qualified leads, quotations in progress. To respect the objective at the end of the quarter, watch the leading ones, because the lagging ones arrive when you can no longer act.
7. It gives qualitative and quantitative answers. Ask ten lost customers why they did not sign, and only then count how many times each reason appears.
8. It distinguishes exploring from reporting. Exploratory metrics search for what you do not know you do not know; reporting metrics keep the ordinary under control.
9. It distinguishes causes from symptoms. A cold can be the cause of a malaise, or the symptom of an influenza that is arriving. Customers who receive a demonstration sign more: but is the demonstration the cause, or does the salesperson give it only to those already convinced? Before making the demonstration mandatory, or removing it, verify which of the two causes the other.
10. It avoids oversizing the metrics. The more numbers you watch, the fewer you use: attention disperses. Keep only those indispensable in the stage you are in, one or a few, and change them when the stage changes. It is the idea of the One Metric That Matters.

Applied by hand, these rules are powerful, but they require hours. And here the real problem is born.

It would be “beautiful” to have two hundred hours a day

It would be “beautiful” to have two hundred hours a day. We have twenty-four, of which eight must be dedicated to work, and this does not allow it. For this purpose we need the technologies of intelligence.

In my experience, when I link artificial intelligence to data science (the discipline that obtains decisions from numbers), but above all to values, to the heritage of knowledge and to internal strategies, the companies that make this step become much faster.

From it is born something that I define as proprietary algorithmic selling, an entry in the glossary of Enrico Giubertoni’s terms. Proprietary, because it is the property of the company.

Selling by hand and proprietary algorithmic selling

	Selling by hand	Proprietary algorithmic selling
How the rules are applied	With spreadsheets and meetings	With a system trained on the data and knowledge of the company
Times	Hours that commercial work does not leave	Much faster
Whose the system is	Whoever compiles it	The company's
Who decides	Whoever has the time	Always the decision maker

The question is this: do we prefer to remain in the era before AI and say “it would be beautiful to have two hundred hours a day”, or do we train and optimise our existing processes, maintaining and strengthening our decisive thinking and cognitive factor?

Proprietary algorithmic selling: slow branches and fast branches

It is a sales algorithm (a procedure that a system executes following rules we taught it) able to distinguish the slow branches from the fast branches of your portfolio.

Slow branches and fast branches of the portfolio

If you need	The system points the salespeople to	What characterises these customers
Sales volumes, that is cash in the short term	Customers who close early	Short cycles, close decision makers, simple offers
More revenue per contract	Long-cycle customers	High value, more people involved, more deliberation

The system is dynamic, because the indications change with the objective, and it is automatic, because it recalculates by itself. Attention, however: automatic means accountable, never delegated.

Automatic means accountable, never delegated

One of the legitimate fears of those who approach artificial intelligence is this:

But how can I delegate a decision process to a machine? And above all: all this makes one feel, rightly, diminished.

The answer is clear: not only you cannot, you must not. A machine is a machine. We are speaking of technologies of intelligence, and they are technologies: the power of decision must always remain with us. Delegating a process to a machine means losing the meaning of that process and, consequently, the control over it. It must not be done.

Whoever photographs with a high-end camera knows the program mode. It is automatic: it chooses time and aperture, removes the ballast, brings you to a result that you can read and decode. But you are the one who chooses, and you have three possibilities:

• use the preset choice;
• refuse it, because you have in mind something more effective;
• modify it.

Nobody removed the camera from your hands. With artificial intelligence in sales the same principle applies. If an important commercial decision has no name and surname beside it, automation has already overtaken what it should have served.

The principle has a name: human in the loop (the human being inside the decision circuit). I prefer, however, to speak of human on the loop: the human being at the top of the decision circuit, and not only involved as a thinking mind in one of its steps. The difference matters, because otherwise artificial intelligence and the tools remain without supervision.

The Golden Rule of AI

1. 20% tool. Artificial intelligence: it serves to execute.
2. 30% data. Theoretical and pragmatic: it serves to understand.
3. 50% knowledge. The company’s: it serves to decide well.

Why does the tool weigh 20%? Because the real added value of AI is not automation. It is how much we manage to insert into AI of the theoretical rules of the best of all possible worlds, those we would apply if we had two hundred hours a day.

Competitiveness, in the twenty-first century, is not played mainly on the best AI. It is played on how much corporate culture you possess to integrate into the decision system of the AI, in order to train it to give precise answers, aimed exclusively at:

• your internal data;
• the processes you already have;
• the exceptions that your salespeople know in depth and apply with full knowledge of the facts.

AI is not magic. It works well only where it finds an orderly heritage.

My professional approach

What you have read is a small taste of my professional approach, the fruit of my method and of my work alongside the sales teams of companies. You find it in full on the page AI Transition, the training on your own processes and your own data. There you find:

• the three phases of the programme: assessment, design of the intervention and delivery, with laboratories on the real processes of each department;
• the outcomes (benefits) role by role: for sales, B2B proposals and answers to RFQs (the requests for quotation from customers) prepared in shorter times, without losing personalisation;
• the possible funding through interprofessional funds (the funds that finance the training of employees), where applicable in the company’s country;
• the programme to download, to read the path calmly and share it inside the company;
• the form to tell me your priorities, at the bottom of the page.

A concrete case study on the sales force is in the article AI Sales Enablement (available in Italian).

Share your priorities

Now I hand the floor to you. Tell me your priorities and your objectives, including the urgent ones that you must reach, and I will give you a targeted answer within one working day.

Share your priorities