How to increase sales: fix the leaks first

Most advice about how to increase sales starts by telling you to make more of something. More leads, more calls, more content, more rows in the CRM. It is the easiest advice to give and the most expensive to follow, because it assumes the hole is at the top of the funnel.

Usually it isn't. Usually the interest was already there, and it leaked out somewhere between the inquiry landing and the proposal going quiet.

I run growth at LiveDocument myself, which means I do the outbound, write the follow-ups, and then sit with the silence afterwards. Before this I was doing go-to-market for a data enrichment SaaS and running growth at a lead-gen company, so I have sent a genuinely embarrassing number of decks and proposals into inboxes and spent a lot of weekends guessing whether anyone read past page two. The pattern barely changes between companies. Almost nothing dies because nobody was interested. Plenty dies because interest wasn't answered fast enough, wasn't qualified properly, or wasn't given anything that could survive being forwarded to someone who never met you.

So this is the version I'd give someone who wants more revenue this quarter and no budget for more leads.

Deals die after the first touch, not before it

A lead can show real interest and still disappear. The first reply lands late. The follow-up adds nothing. The proposal assumes one person can persuade everyone else in the building. Each of those looks minor on its own, which is exactly why activity keeps climbing while revenue sits still.

Speed is the one with the clearest evidence behind it, though the evidence is messier than the numbers suggest. Some vendors report that responding within five minutes makes you around 21 times more likely to qualify a lead, against an average B2B response time of roughly 42 hours, according to summaries of lead-response research. Those figures come from vendor roundups rather than a primary study, and I've never found the original behind the 21x, so treat the multiple as a shape rather than a measurement. The direction holds either way. By tomorrow morning your buyer has spoken to somebody who answered while the problem was still annoying them.

The buyer isn't waiting for another sales call

Most of the evaluation now happens without you in the room. Only 17% of the B2B buying journey is spent with vendor sales reps, down from roughly 40% in 2017, and the average enterprise buying committee has grown to about 11 stakeholders, reaching 20 on the most complex deals. Both figures are Gartner's, collected in this B2B buying-journey research. If your proposal, pricing, security page or rollout plan leaves a gap, your buyer has to fill it themselves, in a meeting you're not invited to.

That internal meeting is where deals actually stall. Your champion is sold. Finance is looking at the number. Operations wants to know who does the work. Legal is reading a clause you've never thought about. Gartner's 2024 numbers put the cost of that crowd at roughly 10 percentage points of purchase probability for every extra stakeholder in the room, which is a brutal way of saying every person you don't equip is working against you.

Your CRM will not tell you any of this. It records a meeting, a stage, and a proposal date. It doesn't know whether the decision-maker ever reached the pricing section, whether your champion reopened the deck the night before the internal review, or whether the finance lead opened it at all.

The three places revenue leaks

Speed. Someone puts their hand up. You reply the next working day. Somebody else replied in twenty minutes and framed the problem first.

Discovery. You hear a surface-level problem, nod along, send the standard deck. At the pricing stage you find out the real buying criteria never came up.

Consensus. Your contact likes it. Nobody else has heard of you, and your contact is now selling your product internally using a document you wrote for them, not for the people who can kill it.

Lead generation still matters. The mistake is buying more volume before you've fixed the handoffs, because more leads through the same three holes just costs more per lost deal.

Discovery should end in one sentence you can quote back

The point of a discovery call isn't collecting pain points. It's leaving with a sentence in the buyer's own words that you can repeat to them later and have them agree with.

Something like: "the issue isn't sharing files, it's that we can't tell which of the four people on the client side actually read the scope before they signed off."

That sentence does three jobs. It proves you listened. It becomes the first line of your proposal. And when your champion forwards that proposal to someone who wasn't on the call, it's the part that makes the document make sense without you standing next to it.

Analysis of recorded sales calls suggests win rates peak somewhere around 11 to 14 discovery questions, according to sales discovery guidance from Perspective. I'd hold that loosely. The count matters far less than whether you left with the sentence.

Start with context, not with your product

Early questions establish what happens today, before you introduce anything. What does the team do now? What changed that made this worth solving? What breaks when the current process breaks? Who spends their week working around it? What would have to be different for this to count as solved?

Don't jump to pricing while the buyer is still describing symptoms. A low number won't rescue a proposal aimed at the wrong problem.

Quantify the impact without running an audit

You don't need a financial model. You need enough detail to tell a real initiative apart from someone browsing. Which teams are affected, what gets delayed, what risk the current approach creates, what happens if they do nothing for another six months.

That last question is the one people skip because it feels rude. It's the one that tells you whether you have a deal or a research project.

Map the buying process before you send anything

Who influences this, who owns the budget, what each of them will judge it on. Then the boring end: approvals, procurement, security review, legal. Ask it on the call and it's a normal question. Ask it in week six and it sounds like panic.

I'd also say this plainly, because it took me too long to learn: the call where the buyer talks most and you say almost nothing clever is usually the one that closes.

Your proposal is selling for you when you're not in the room

A static document makes the reader do all the interpretation. That's fine when the decision is simple and one person owns it. It falls apart the moment there are scope assumptions, two pricing options, technical dependencies, or terms that four people will read four different ways.

Think about what each of those readers needs. A consultant sending a strategy report has a client who reads the executive summary, skims the recommendations, and never reaches the implementation section that explains what to do on Monday. A founder sending an investor deck has partners reviewing the market slide and the traction slide separately, on different days, without the founder there to explain why the market number is what it is. A rep sending pricing has a buyer who needs to know which assumptions move the figure.

In all three cases the fix is the same: put the explanation next to the document instead of in the covering email, because the email doesn't get forwarded and the attachment always does.

Narrate the decisions, not the slides

Don't record yourself reading the page out loud. Record the parts that change how the page gets read. Why this approach and not the obvious cheaper one. Which assumption would move the price. What this section means for the person in finance who's about to open it cold. What you want to happen next.

Keep it specific to the recipient. "This page shows our standard process" is filler. "This page is the handoff between your ops and finance teams, which is the thing you said was costing you a week" gives the page a job.

This is the problem I ended up building LiveDocument around, so read this with the appropriate suspicion. You record a short walkthrough, attach it to the PDF or image, and send the whole thing as one link, with clickable highlights that jump the video straight to the section the reader is looking at. It's a video walkthrough attached to the document itself, not a Loom link sitting in a separate tab hoping someone connects the two. It's also not a data room, not e-signature, and not storage. Links expire and can be revoked, but there's no watermarking and no NDA gating, so if that's your requirement, this isn't your tool.

Pick the documents worth the effort

Start with the ones that reliably generate a clarification email or decide something meaningful. Proposals, pricing, implementation plans, service reports, investor decks. Don't turn every routine attachment into a production.

The test is simple. Could someone who missed the call understand the recommendation, the reasoning, and the next decision from the link alone? If yes, you're done. It doesn't need to be polished.

Follow-up needs a reason, not a reminder

Follow-up fails when persistence gets confused with repetition. Four rounds of "just checking in" is not a cadence. It asks the buyer to do work and gives them nothing in return.

Most articles on this subject open with the same two benchmarks about how many follow-ups a sale needs and how many reps quit after one. I've quoted them myself in a past life. Then I went looking for the studies behind them, traced both back to blog posts citing other blog posts, and could not find a primary source for either, so I've stopped repeating them. What survives without a citation is the shape, and it matches what I see every week: the early touches carry most of the value, so front-load the effort instead of grinding out touch seven.

Give the first follow-up a job

The first one recaps the sentence you got in discovery, corrects anything you misheard, and points at the section of the proposal that answers it. Write it so it can be forwarded, because it will be.

Later touches need their own reason to exist. The second answers the question they haven't asked yet, usually implementation or what happens in month two. The third offers something concrete: a short call on one unresolved point, or the reference they'd want before committing. The fourth is the direct one: "Has this dropped down the list, and should I stop chasing you?" That last message gets replies at a rate that surprises people, mostly because it's the only one in the sequence that costs nothing to answer.

Personalization helps, but not the kind that means merging a first name into a template. The reported gain, personalized emails performing around six times better than generic ones, comes from tailoring to role, concern and stage, as summarized in this B2B follow-up practice research. Merge tags are not personalization. They're mail merge with better PR.

Escalate, don't just repeat

Change the channel or the value as the deal moves, rather than extending the same sequence. Written recap, then a useful clarification, then a short call tied to one unresolved question, then the note that makes it easy to close or park.

Then stop. Four touches over a couple of weeks is what keeps this week's outbound from competing with three weeks of unfinished chasing, which is the failure mode I've written up in the sales prospecting plan and the process that runs off it.

Engagement data tells you when, not whether

A calendar tells you when to send. Engagement data tells you what the message should be about. Confuse the two and you send irrelevant outreach at exactly the moment the buyer was looking at your pricing.

PDF tracking software is genuinely useful here and I use it weekly. It's also routinely over-read. A reopen at 11pm on a Tuesday isn't buying intent. It might be your champion prepping for a meeting, someone clearing an old tab, or the finance lead opening it for the first time.

Read signals as questions, not verdicts

A prospect reopening the pricing page might be comparing options, might be preparing an internal argument, might be stuck on an assumption you didn't explain. All three call for a question, not a close.

Signal	What it usually means	What to send
Fresh open after a quiet stretch	Something restarted the decision, or a meeting is coming	A short note tied to the section that matches where the deal actually is
Repeated views of the pricing page	Scope, terms or a comparison needs clarifying	The assumptions behind the number, and a question about which option is under review
Long attention on one technical page	Feasibility or risk is being evaluated	A specific answer, or a short call with whoever can give one
Several people viewing in one afternoon	It may have been forwarded, so alignment could be starting	Ask your champion who else is looking, then send something forwardable for them
Downloaded, then nothing	It's gone offline and intent is unknown	Ask whether it answered the main question, and offer to clarify one section

The action should match the signal and the gap in the deal. Don't chase every alert. Knowing someone opened your document is not the same as knowing they understood it, and treating the analytics as if it were is how reps end up chasing ghosts with real enthusiasm.

The weekly version of all this

None of it survives without a slot in the calendar, so here's the small version.

Monday, look at every open deal and ask which one has attention but no consensus. That's the one to work. Through the week, answer new inquiries the same day, and after every discovery call write down the buyer's sentence before you write anything else. When you send a proposal, record the two-minute explanation with it. Friday, twenty minutes: what got opened, what went quiet, who leaves the sequence.

No new leads anywhere in that, and it's most of what's moved our own numbers. The version aimed at sales teams rather than founders doing six jobs at once is on the sales use case page.

I built LiveDocument because I got tired of sending careful work into inboxes and guessing. If that's familiar, it's at livedocument.com.