FR
EN
Article
The YBA.ai Team · October 2026

The customer needs an answer. The rep is looking for it across three tools. The deal slips. Another app will not necessarily solve the problem.
272 days. That is the average B2B buying journey measured by Dreamdata in its 2026 report, from the first recorded interaction to a closed-won deal. And 81% of that journey happens before the account enters the sales pipeline [1].
That does not mean following up with a prospect for seven months. It shows how much time can pass between a company first discovering your name and buying from you.
During that time, it researches, compares and discusses internally. On the sales side, we cannot simply wait. We need to understand why the company is interested now, who is involved in the decision and what it needs to move forward.
But we also need time to do that work.
Your contact is convinced. Their company is not there yet.
Imagine selling software to a manufacturing company. The demo went well. The operations director sees the value. The rep has sent the proposal. Then replies start to slow down.
In the pipeline review, their director asks: “Have you followed up?” Yes. Twice.
But the finance director has not approved the budget. The IT director is waiting for answers about integration. Site managers want to know who will train their teams. The rep’s contact needs to convince all of them.
The same Dreamdata report counts an average of 10 stakeholders and 88 touchpoints in the journeys studied [1]. That does not mean 88 emails. These are interactions across several channels, around a decision that involves more than one person.
A convinced contact is not enough to close a deal. The company needs to be able to buy.
In our example, sending everyone the same presentation is not enough. Finance needs the costs and expected return. IT needs to understand integration. Site teams need to know how much work deployment will involve. The rep needs to prepare those answers with their contact, without going behind their back.
The proposal is out. What is holding up the close?
“Demo completed”, “proposal sent”, “follow-up scheduled”: the CRM describes what the rep has done. It does not necessarily say what the customer still needs.
If the budget has been postponed, we need to understand when the project can resume. If an integration answer is missing, we need to provide it. If our contact does not know how to make the case internally, we need to help them present it.
These are three different situations. Scheduling three follow-ups does not resolve them.
On 27 October, Dreamdata published an interview with Louis Grenier on buying triggers [2]. His advice: ask recent customers what made them act. A problem can last for years. A new contract or a change in leadership can suddenly make it urgent.
That attention needs to continue after the first meeting. What has changed at the customer’s company? Which question is still open? What next step have we actually agreed with them?
How many tools do you open to answer one customer?

To find those answers, the rep opens the CRM, searches for the latest email, checks a shared folder and looks at the tracking spreadsheet. Then they ask pre-sales whether anyone has handled the question.
We have added one tool for contacts, another for emails, another for quotes. But someone still has to move the lists, check for duplicates and find the right version of the document.
A new SDR or account executive has to learn where to look, what to copy and which file to trust. The team spends time keeping its tools working while the customer waits for an answer.
In a recent meeting with a B2B auction platform, the team was happy with its campaigns. But lists were being extracted and imported manually into the sending tool. The needs raised included reducing that handling and finding more relevant contacts.
That conversation gets to the point: why replace a campaign that works if the time is being lost in preparation?
To unblock deals, we review the sales with the reps
This is where we start at YBA. We examine won, lost and stalled deals with the reps and their director. We go through emails, meeting notes and CRM stages to understand where the sale moved forward and where it stopped.
The rep adds what the tools do not tell us. Pre-sales, marketing or finance join when the issue involves them. We check the explanations together before deciding what to change.
The founding team brings more than twenty years of sales experience, including at Oracle and Salesforce. That experience helps us distinguish between a targeting problem, a missing approval at the customer and a delay between teams on the seller’s side.
This is the Sales Operating Model diagnostic: reviewing how the company organises selling, from prospecting through to closing. The aim is to identify what works and fix what is wasting time.
Selling more does not always take another tool
Back to our software deal. If the customer is waiting for an integration answer, the first improvement is to agree who will respond, by when and where to find previously checked answers.
The rep can then send the document the customer needs and agree the next step with their contact. The CRM keeps the outstanding question and its owner. A spreadsheet that repeats this information without another purpose can go.
With prospecting lists, we look for the same opportunities to reduce repeated exports, imports and checks, while keeping useful campaigns. Sometimes, connecting two tools properly is enough to avoid several manual steps.
We test these changes with one team or part of the pipeline. We compare time spent, response times and the steps customers complete before extending them.
AI agents should free up time to sell
Once this work has been reviewed, AI agents handle some of the preparation in connected tools: researching the right people, preparing messages and finding documents needed for a deal.
For a stalled deal, the agent analyses the available conversations and prepares an assessment of what is missing. The rep checks it, adds information absent from the CRM and chooses what to do next. The prepared message needs to address the customer’s situation.
The team keeps the conversation, the negotiation and the close. We also help reps adopt the new way of working. An assistant they have to feed manually all day would not save them much time.
Doing more with less should show up in the sales

Across client deployments measured before and after, per agent, we have recorded results including an 8% revenue increase and a 91% reduction in manual work [3]. These are results from the deployments concerned, not an identical promise for every team.
Reducing manual work does not reduce total costs by the same amount. Subscriptions, deployment and time spent on each account all count when checking what selling actually costs.
The result we want is concrete: fewer manual steps, faster answers and more time to move deals forward.
A 272-day buying journey does not justify 272 days of waiting. When a sale drags on, the useful question is not just “Have you followed up?” It is also: “What does the customer need from us, and why have they not received it yet?”
At YBA, we review these deals with your teams to identify where time is being lost and what we can simplify in your sales process.
To discuss this, email us at yba@yba.ai.
Sources
[1] Dreamdata, 2026 LinkedIn Ads B2B Benchmarks Report, official carousel, pages 2–3. 2025 data: 272 days from first recorded touch to closed won; 81% before entering the pipeline; 10 stakeholders and 88 touchpoints. The report covers Dreamdata customers; its page describes more than 66 million sessions and 3.5 million journeys. These averages do not represent all B2B sales.
[2] Stephanie Dean, “The 4-Step Framework for How B2B Brands Get Chosen”, Dreamdata, 27 October 2026. Interview with Louis Grenier on buying triggers, separate from the report data.
[3] YBA before-and-after measurements, per agent in client deployments, included in our materials dated 27 October 2026. Results reported by YBA; sample sizes and measurement periods are not specified in this source. The reduction in manual work does not represent an equivalent reduction in total costs.