The 7 sales process steps, and the data each step runs on

Every guide to the sales process steps gives you the same seven boxes: prospecting, preparation, approach, presentation, handling objections, closing, follow-up. The model is a century old, it is genuinely useful, and it is where almost every article stops. What none of them tell you is that four of those seven steps are not skill problems at all. They are data problems, and a rep with a perfect script will still stall on them if the row in front of them is wrong.

This guide walks the seven sales process steps in order, and for each one names the specific thing that has to be true about your data for that step to work. That is the part that decides whether the process runs or quietly leaks.

The sales process steps, in one table

The classic model is linear on paper and circular in practice. Here is the whole thing before we take it apart, with the failure that is most common at each stage.

Read the failure column again. Steps 1, 2 and 6 fail because of something you knew or did not know about the account. That is the thread this guide pulls.

Sales process step 1: prospecting is a data problem

Prospecting is where the process is won or lost, and it is the step most often described in motivational terms rather than operational ones. "Build a list of ideal customers" is not an instruction. The instruction is: define the firmographic filters that describe an account worth your time, pull every company that matches, find the person in the relevant function, and get a contact point that resolves.

Three things break here, in order of how expensive they are. The first is targeting: a list assembled from whoever was easiest to export is not a list of your ideal customers, it is a list of the well indexed. The second is completeness: a company row with no named contact is not a prospect, it is a logo. The third is deliverability: an email address that bounces did not just fail, it damaged the sending domain that the rest of your list depends on.

The mechanics of getting this right are a subject of their own, and the step-by-step version with the real cost per thousand rows lives in the guide on how to build a prospect list. The short version: pick the filters first, pull the companies, then resolve people inside them, and verify before anyone sends anything.

This is also the step where tooling changes the shape of the work rather than just the speed. Pulling a few hundred matching companies and finding the right person inside each one is a batch operation, not a browsing session. Derrick runs it wherever your list already lives: the sidebar in Google Sheets for a column of rows, the MCP server when you are asking a question inside an AI client like Claude, and the REST API when it is a step in an automated workflow. Enrich Companies costs 1 credit per company and works on the free plan, which is 100 credits a month at no cost.

Preparation, and the hours it quietly costs

Preparation is the step that looks free and is not. The classic advice is to research the account before you reach out, which is correct and completely unactionable at volume. Six minutes of manual research per prospect sounds trivial. Across two hundred prospects in a month it is twenty hours, which is a salesperson not selling for half a week.

What the rep actually needs before the first contact is short and specific: what the company does in plain language, roughly how big it is, whether it matches the profile of accounts that already buy, who else sits on the buying committee, and one recent fact that proves you looked. Everything beyond that list is procrastination dressed as diligence.

The distinction that matters is between research a human must do and research a machine should have already done. Headcount, sector, location, technology in use and the shape of the org chart are all attributes, and attributes should arrive with the row rather than being fetched one browser tab at a time. What stays human is judgement: whether this account has the problem you solve, and what to say about it.

Steps 3 and 4: approach and presentation

The approach is the first real contact, by phone, email or a LinkedIn message. Its only job is to earn a conversation, which is a much smaller ask than earning a sale and is routinely confused with it. An opener that describes your product is a pitch. An opener that names something true and specific about their situation is an approach.

Specificity is where preparation pays back. The difference between "we help companies like yours improve efficiency" and a sentence that references what their team actually does is entirely a function of whether the data arrived with the row. If it did not, the rep improvises, and improvisation at scale reads as a template.

Presentation is the step where most teams over-index on their own product. The demo that converts is not the one that shows the most features, it is the one that walks the buyer through their own workflow with your product inside it. That requires knowing the workflow, which loops back to step 2 again. If your presentation is identical across accounts, you are not presenting, you are broadcasting.

Channel choice belongs here too. A phone conversation and a written approach fail for different reasons and are worth reading about separately, which is what the guide on outbound sales covers in detail.

Steps 5 and 6: objections and closing

Handling objections is the most misunderstood of the sales process steps, because the word "handling" suggests something to be overcome. An objection is information. It tells you either that the value has not landed, that the timing is wrong, or that you are talking to someone who cannot say yes. Only the first is a selling problem.

The third is a data problem wearing a selling costume. "I need to check with the team" is very often not a stall, it is an accurate description of a buying committee you did not map. In B2B the person who takes the first meeting is frequently not the person who signs, and the number of people involved has grown steadily. Mapping that committee before the close rather than discovering it during is covered in the guide on how to sell to decision makers.

Closing, when the previous five steps have been done, is usually undramatic. The theatrical close belongs to a world where the seller controlled the information. Today the buyer has read your pricing page, your documentation and two alternatives before the first call. A close in that world is an agreement on next steps, not a performance.

Follow-up, where the revenue actually is

The seventh step is the one that gets cut when the quarter is tight, and it is the one with the best return. Follow-up covers three distinct jobs that get lumped together: pursuing deals that did not close, onboarding the ones that did, and expanding accounts that are already paying.

Most pipelines lose more revenue to silence after a "not now" than to outright losses. A "not now" is a dated fact, not a verdict, and the thing that makes it actionable later is having recorded when to return and what changed since. That is a CRM hygiene question, and CRM hygiene is a data question: a record whose contact left the company eighteen months ago is not a follow-up opportunity, it is a wrong number.

Contact data decays continuously as people change roles. A follow-up sequence built on a list captured a year ago will quietly send a meaningful share of its volume to addresses that no longer exist, which is both a wasted touch and a deliverability cost. Re-verifying before a re-engagement campaign is the cheapest step in the entire process.

The data layer under the sales process steps

Here is the same seven steps, mapped to the specific attribute that has to be present for the step to work. This is the table the other guides do not print.

Notice that steps 3 and 7 depend on the same thing: a contact point that resolves today. Finding one is a discrete job with a price. Email Finder costs 5 credits per email and only charges when a result is found. Phone Finder costs 150 credits per phone number on the same find-only basis. Both are on the paid plans.

Adapting the seven steps to your own cycle

The seven-step model is a scaffold, not a script. A self-serve product with a fourteen day trial and an enterprise deal with a procurement review share the same skeleton and almost nothing else. Three adjustments cover most cases.

• Collapse steps when the deal is small. Below a certain contract value, approach and presentation are the same message. Running six formal stages on a deal worth a few hundred euros a year costs more than the deal returns.
• Split step 6 when a committee is involved. Enterprise closing is rarely one event. Legal, security and procurement each behave like their own mini-cycle with their own objections.
• Move step 2 upstream when volume is high. At scale, preparation stops being something a rep does per account and becomes something the list arrives with. This is the single highest-leverage change most teams can make.

The one thing not to adjust is the order. Every attempt to skip prospecting rigour and make it up later with better closing produces the same result: a full pipeline of accounts that were never going to buy, and a team that concludes it has a closing problem.

Whether your motion is inbound or outbound changes which steps carry the weight, and the comparison is worked through in inbound vs outbound marketing. If most of your pipeline raises its hand first, the guide on inbound sales is the closer fit.

Five ways the sales process steps break

• Measuring activity instead of progression. Calls made is an input. Deals moving from step 3 to step 4 is the number that tells you whether the process works.
• Treating the process as a reporting structure. If the stages exist so the forecast has columns, reps will update the CRM to look right rather than to be right.
• Blaming closing for a prospecting failure. Deals that die at step 6 were frequently mis-qualified at step 1. The symptom and the cause are five steps apart.
• Leaving preparation to individual discipline. Anything that depends on every rep choosing to do it consistently will be done inconsistently.
• Never revisiting the ICP. The definition that was right two years ago quietly stops matching who actually buys, and nobody notices because the pipeline still fills.

Where to start

Pick the earliest step that is leaking and fix that one. Teams tend to work on the last step that hurt, which is usually closing, and closing is usually a symptom. If your reply rates are low and your bounce rate is not, the leak is targeting. If your bounce rate is high, the leak is the contact data. If demos feel generic, the leak is preparation.

Two of the seven steps can be improved this week without changing anything organisational, because they are supply problems rather than behaviour problems: the completeness of the list, and the accuracy of the contact point. Start on the free plan with 100 credits a month and run one week of prospecting through an enriched list before you change a single script. The full catalogue of what can be resolved per row is on the features page.

We publish one email every 2 weeks with playbooks like this one and the data behind them. The capture form below this article is where it goes out.