Ishikawa diagram: how to find the real cause of a sales decline

Why sales teams get stuck at the diagnosis stage

A typical scene: the monthly report shows that the number of quote requests dropped by a dozen or so percent, and the sales team cannot point to a single cause. Salespeople talk about lead quality, marketing about clients' overly long decision cycles, and leadership about competitors cutting prices. Each of these answers is plausible, none of them has been verified.

In B2B companies this moment is especially costly, because purchasing decisions take longer and a single client weighs more than in e-commerce. By the time the team manages to test one hypothesis, a quarter has passed. That is why a tool is needed that gathers all possible causes in one place and makes it possible to organize them. The Ishikawa diagram, also called the fishbone diagram, does exactly that.

At Growth Group we work with companies across the region and beyond. Over 16 years, more than 60 clients have passed through our projects, and we have generated PLN 160 million in profit for clients. In practice the biggest problem is a lack of structure in which data can be arranged. The Ishikawa diagram provides exactly that structure.

What the Ishikawa diagram is and when it truly pays off

The Ishikawa diagram is a graphical method of cause-and-effect analysis. On the right side sits the problem (the effect), and on the left lines run into it representing the main categories of causes. Each of these lines branches into detailed factors. The whole thing resembles a fish skeleton, hence the colloquial name: fishbone.

The method originated in Japanese industry, but in B2B marketing it works just as well, because problems here also have many overlapping sources. The diagram is worth using in three situations:

• The problem is complex and recurring. Falling conversion, rising customer acquisition cost, high turnover in the sales department. If the problem comes back every quarter, it is worth diagnosing it properly once.
• The team has different hypotheses and cannot decide. The diagram shows all hypotheses side by side and makes it possible to assess which ones have the greatest potential.
• There is not enough data for a quick decision. Instead of guessing, the team gathers a list of factors and plans which ones to measure first.

The diagram does not replace data. It organizes the questions for which data still needs to be collected. This is an important distinction, because many teams treat it as a tool for quick conclusions, and end up with a nice drawing and zero decisions.

How to build the diagram step by step

Below is a process that works in teams of three to twelve people. The whole thing takes 60 to 90 minutes if the problem is well defined.

Step 1: write down the problem concretely

At the end of the axis on the right side you place the effect. Instead of "poor sales results", write "an 18 percent drop in the number of closed deals in the second quarter". The problem must be measurable and limited in time, otherwise the diagram will scatter into ten different topics.

Step 2: choose the main categories

The classic set is man, machine, method, material, measurement and environment. In B2B marketing an adaptation works better: people, process, technology, content, market, measurement. Choose four to six categories. More dilutes attention.

Step 3: gather factors in each category

This is the brainstorming stage. There is one rule: you do not evaluate, you just write down. In the "people" category you might see salespeople's competencies, turnover, the way they are compensated, the quality of training. In the "process" category the length of the sales cycle, the number of stages, the moment marketing enters. Each factor is a separate bone.

Step 4: assess impact and ease of verification

Rate every factor on two dimensions: how much impact it could have on the problem, and how easy it is to verify whether it is actually at play. Factors with high impact and easy verification go to the top of the list. Those with high impact but hard to verify need a separate research plan.

Step 5: assign owners and deadlines

A diagram without owners is a wish list. Every factor in the first priority group should have a person responsible and a date by which the data will be collected. Without that, the meeting ends with enthusiasm that evaporates by next Monday.

Fishbone example: conversion decline at a B2B company

A manufacturing company approached us with a problem: the number of quote requests stayed flat, but conversion to contract dropped by over ten percent within two quarters. The team had three hypotheses: lower lead quality, a longer client decision cycle, weaker sales performance.

We built a fishbone diagram across six categories. After an hour of work the list of factors ran to more than thirty items. Three turned out to be key:

• Process: marketing had stopped qualifying leads before handing them to sales. Salespeople were receiving inquiries from companies with no budget and no decision-making authority.
• Content: the case studies section for two key industries had disappeared from the website. Clients had nothing to confirm that the company understood their specific needs.
• Measurement: the team was tracking the number of leads rather than the number of qualified leads. The metric was rising, quality was falling, and nobody saw it.

None of these causes was visible in the weekly sales report. The fishbone diagram forced the question of what happens between lead generation and handover to the salesperson. That stage turned out to be the bottleneck.

In similar situations a broader marketing audit is often needed, covering not just the sales funnel but also positioning, communication, and customer acquisition channels. A fishbone diagram is a good starting point, but it does not answer whether the company is even speaking to the right audience.

Cause and effect analysis: from diagram to decision

The diagram itself is only half the work. Cause and effect analysis ends the moment the team decides what to change. Below are three principles that in practice determine whether the analysis delivers results.

Principle one: one change at a time

If the team introduces five fixes at once, a month later nobody knows which one worked. Choose one, at most two factors with the highest priority and test them in isolation. In B2B, cycles are long, so patience matters here as much as an accurate diagnosis.

Second principle: measure what you change

Every change must have a metric assigned to it. If you improve lead qualification, measure the percentage of qualified leads, not the number of inquiries. If you shorten the sales cycle, measure the time from first contact to decision. A metric that is not tied to the change will show nothing.

Third principle: return to the diagram after a quarter

The Ishikawa diagram is not a document for a single campaign. Come back to it every quarter, add new factors, cross out those that have been ruled out. Over time it becomes a map of knowledge about the company, not a one-off exercise.

In companies that run several product lines or sell to different segments, it is worth building a separate diagram for each area. A shared diagram usually blurs into generalities and stops being useful. If the company operates in a model where marketing and sales have different goals, it is good to first establish a shared B2B marketing strategy, and only then diagnose individual problems.

The most common mistakes when working with the diagram

After more than a dozen years of working with sales teams, we see the same missteps. It is worth knowing them before you sit down for your first session.

• A problem that is too general. "We have too few customers" is not a problem, it is a state that needs to be broken down into causes. The diagram needs specifics: a drop, a rise, a deviation from the norm.
• Too many categories.Six is the maximum. With eight, participants lose track of where to write what.
• Mixing causes with effects.If a factor is itself the consequence of another, it does not belong in the diagram. That is a signal the problem needs to be split into two separate ones.
• No verifying data.A diagram without a measurement plan is a list of hypotheses. Hypotheses are only useful when someone tests them.
• Working alone.A diagram built by one person reflects that person's limitations. The value appears when people from different departments sit at the table.

The last mistake is especially common in companies where marketing and sales operate in separate silos. An Ishikawa diagram forces a conversation, because the "process" and "people" categories concern both teams at once.

When an Ishikawa diagram is not enough

An Ishikawa diagram is a diagnostic tool. It will not replace the decision about which customer group the company wants to invest in, nor will it help if the problem is a lack of coherent strategy. Those situations call for a broader view: market analysis, positioning, offer structure, reach channels.

It is also worth remembering that the diagram does not work well in teams that have no access to data. If a company measures nothing beyond revenue, the diagram turns into a brainstorm with no reference point. In that case the first step is implementing basic measurement.

For companies that want to go through such a diagnosis with external support, we offer two working models: full agency service and an interim marketing manager paid on results. In both cases we start by understanding the problem, not with a ready-made list of actions.