Systematizing Sales: What Five Documented Cases from Research Show

Why is it worth looking at other companies’ cases?

There is no shortage of advice on sales. Most of it describes how things should be. You rarely learn what a company actually did, what the situation was beforehand, and what could be demonstrated. That is exactly what case studies published in academic journals provide. They state method, data basis, and limitations.

This article presents five such cases. It reports only what the authors state in their published papers and puts it into context for small and mid-sized businesses. The cases come from different countries and industries. None is a recipe. Together, however, they show commonalities that can be applied to a business with ten or fifty employees. The guiding question is therefore: what did companies that demonstrably systematized their sales actually do? To answer it, we first give an overview of the cases, then go through them one by one, and finally summarize what they have in common in a pattern of four building blocks.

Systematizing starts with a number you did not know before, and rarely with a piece of software.

Which cases are documented? An overview

The cases differ greatly in how robust they are. Case 5 rests on a field experiment with a control group. Case 1 measures baseline values and sets targets. Case 3 comes from the originator of the model studied. When reading, these differences matter more than the individual percentages.

Case 1: How does an IT consulting firm organize its presales?

Surdziel and Gdowska (2026) accompanied a medium-sized Polish IT consulting firm. Presales there ran informally and, as the authors describe it, kept drifting away from the intended workflow. This limited coordination, transparency, and growth. The researchers collected data in twelve interviews and two workshops with sales, technical staff, finance, and delivery. The existing workflow was recorded as a model and examined with value stream mapping and root cause analysis.

No prior performance data existed. The baseline values were therefore estimated from several sources. Only then could targets be set.

The measures were plain. The company established a dedicated presales department, defined roles more explicitly, and systematically involved technical expertise from the start. Simple aids were added: process models, a form for project barriers, and a form for analyzing won and lost bids.

As early outcomes, the authors name reduced process drift, better cross-functional collaboration, and more visibility for management. Their conclusion: even in businesses with low process maturity, incremental, lightweight interventions yield meaningful improvements, and they do so without advanced technology. This matches the principle from the article on why the process comes before the tool.

Case 2: Why do sales targets fail in practice?

Morelli and Braganza (2012) examined a pharmaceutical company in Spain in an in-depth case study. They classify it as mid-size and call it Alpha. It employs a sales force of 250. Their starting point: goal setting is regarded as one of the main processes in managing sales forces, yet developing effective sales quotas proves difficult.

The authors interviewed 16 people involved, from the general manager to sales representatives, and analyzed about 200 pages of transcripts. They found eleven barriers. Among them are an unrealistic overall forecast, the lack of a person who owns target setting as a whole, no participation by the salespeople, a formula nobody can see through, missing information, and no fixed channels through which knowledge flows back from the field. Where data was lacking, the revenue target was divided by the number of salespeople without looking at the territories. From earlier research the authors adopt three general reasons why quotas are difficult: complex systems, very granular data, and too little attention to implementation. They name their main limitation themselves: a single company was studied.

For a smaller business, this carries a reassuring message. If even a company with its own sales controlling and data analysis gets stuck on missing ownership and an opaque formula, a simple calculation on one page is not a makeshift. It is the more robust solution. What such a calculation looks like is shown in the article on how to plan sales targets.

Cases 3 and 4: What does selling in fixed phases achieve?

Five industrial companies in Latin America

Urdaneta-Ocando (2025) describes a model that structures the sales of industrial companies into five sequential phases: identify the market, segment customers, design the value proposition, sell consultatively, follow up after the sale. The analysis covered the literature from 2020 to 2025, five corporate reports, and interviews with commercial managers.

According to the abstract, the companies focused on strategic sectors and thereby reduced the dispersion of their sales efforts. Three achieved segmentation rates above 70 percent. For the conversion of proposals into contracts, improvements in the range of 15 to 25 percent are stated. Two companies completed a shift toward consultative selling and reported a 20 percent increase in key customer retention. Post-sales follow-up developed unevenly. Two firms reached satisfaction levels above 85 percent.

Two caveats belong here. The author evaluates the model he named himself, which is not an independent test. And the figures come from reports and interviews, not from a controlled comparison. What is robust is above all the direction: those who first define whom they work for sell in a more targeted way afterwards. This fits what the article on the four phases of scaling sales describes as a precondition for handing over.

Medtronic: from product to solution

Xie (2026) studies the medical technology manufacturer Medtronic in an early stage of the shift from pure product sales to solutions with a service component, known as servitization. The basis is semi-structured interviews with four key participants as well as corporate websites and scholarly literature. According to the study, the company aims to improve its operations and strengthen its competitive advantage with this approach and a stronger customer focus. As its contribution, the paper names a framework that captures stages and characteristics of such a transformation.

For small businesses, the corporation is no model in terms of scale. What carries over is the idea that a sales operation selling solutions needs a different organization than one selling products: different conversations, closer coordination with service, longer support for the customer. With four interviews, the data basis is narrow.

Case 5: What happens when pricing decisions are systematized?

Karlinsky-Shichor and Netzer (2024) worked with a US metals retailer that mainly supplies aluminum to industrial firms. There is no price list, and every salesperson prices every quote individually. From roughly 140,000 pricing decisions by 17 salespeople over 16 months, the researchers built a model for each salesperson that learns and reapplies that person’s own pricing policy. In a field experiment over eight business days with 2,075 quotes, the salespeople were shown their own model’s price recommendation in real time for randomly selected quotes.

The result: for the quotes with a recommendation, profits were 11 percent above the control group, even though the salespeople followed the recommendation fully or in direction for just under a fifth of the quotes. The authors attribute this to the model evening out fluctuations in the salespeople’s behavior over time. The second half of the finding is remarkable. For unusual or complex quotes, the salespeople themselves generated higher profits than the model. The authors therefore propose a division of labor in which standard cases are allocated to the model and special cases to the human. In a recalculation with the existing data, the expected profit of this hybrid was 7.8 percent above that of the salespeople alone. The study covers a single company, and the model also adopts the lasting peculiarities of each salesperson.

You do not need machine learning for this. The lesson is: rules for the normal case beat daily form, and experience remains irreplaceable for the special case. A written price list with fixed discount limits is the small version of the same idea. Where technology saves time in sales today and where its limits lie is compared in the article on AI in sales.

A counterpart shows that technology does not work by itself. Habel, Hartmann, Wiseman, Ahearne, and Vaid (2025) examined automated lead nurturing systems in 18 interviews and three data studies with 748, 13,525, and 5,040 leads. In the second study, the probability of conversion rose by 23 percentage points on average, while the other two showed no effect on average. The effect depended on the length of the sales cycle, the expected sales volume, and the existing relationship. The authors recommend verifying the effect in your own context.

What do the cases have in common? The Four-Building-Block Pattern

Put the cases side by side and four elements recur. Together they form the Four-Building-Block Pattern, the mental model of this article. It describes what was present in the companies studied when systematization showed an effect, and what was missing when it stalled.

1. Measure

   It starts with a baseline value, even if it is estimated. In case 1, only win rate and response time made targets possible. In case 5, transaction data provided the foundation.

2. Clarify roles

   It is defined who is responsible for what. Case 1 led to more explicitly defined roles and a dedicated department. Case 3 begins with the question of which customers sales works for.

3. Simple aids

   Forms, a price recommendation, a workflow in a few phases. Case 2 shows the flip side: a target formula that nobody in the field could trace became an obstacle itself.

4. Learn from wins and losses

   Results flow back. Case 1 uses a form for won and lost bids. Case 5 returns salespeople’s own pattern to them as a recommendation.

What stands out is what did not come first in any case: buying a large system. Where technology played a role, it built on existing data and clarified workflows. How to get the first measurement done in half an hour is described in the article on sales bottleneck analysis. The rules for the individual handoffs are explored further in the article on how to optimize your sales process.

Case studies also show that order at the core often matters more than expansion. That holds beyond sales, as the story of Lego in 2004 shows.

What you can put into practice right away

Check your own sales operation against the four building blocks. Every item you cannot tick is a possible next step. Choose a single one.

• I know my close rate and the time from inquiry to quote, at least as an estimate.
• For every sales stage it is defined who is responsible.
• There is a written description of which customers we work for and which we do not.
• Prices and discount limits for standard cases are written down.
• Lost quotes receive a reason from a short fixed list.
• Once a month someone looks at these reasons and draws a conclusion.

The Polish case began with two estimated numbers. You can start with that today: estimate your close rate and your average time to quote, write both down, and check them against the actual cases in four weeks.

The companies studied measured, clarified roles, introduced simple tools, and learned from won and lost quotes. In none of the cases did it start with buying a large system. The cases are individual cases and do not prove that the same steps work everywhere. They do show, however, that systematizing starts with a number, not with software.