Use cases. The situation you are actually in, and what we do about it.

Nobody hires a pipeline agency because they want more content. They hire one because something specific has stopped working: the traffic does not convert, the founder is the only closer, the AI answer names a competitor, the account list has no way in. These are the eight situations we get called about, what the first ninety days look like in each, and the fifth of the work that produces most of the result.

The short answerLemniscate Growth is hired for eight recurring situations: traffic without pipeline, founder-led sales at its ceiling, invisibility in AI answers, a partner ecosystem that produces no referrals, a target account list with no route to the buying group, deals that stall after the demo, a new market with no local proof, and marketing that cannot prove what produced the pipeline. In each one we build the middle of the funnel first, the comparison, proof and objection handling a committee needs, and we run only the fifth of the work that moves the pipeline number.

The situations below come from client programs across SaaS, partner ecosystems, crypto infrastructure, supply chain and higher education

CipherBCKNNXAavenir4CE CloudLabsSt. Martinus UniversityIQLECTBangDBITT DigitalQuills.aiSystemSoft TechnologiesJaspreet BindraBonfireGATCOINHong Kong FinTech WeekVentiveURSA CloudNexChange GroupDatTong TradingPhantom TechHidden BrainsSolvedexNavan AIINTEGProvIQTradeDog

The house thesis

Most agencies sell the top of the funnel. The money sits in the middle.

This is the argument behind every page in this section, so it is worth making properly rather than as a slogan.

Where the decision happens

Where funnel work is usually bought, against the stage where the decision is made 01 Top of funnelImpressions, rankings,followers, downloads02 Middle of funnelComparison, proof, objections,buying group03 Bottom of funnelMeetings, demos, securityreview, procurementCheap to produce, and worth less for itAnyone can publish it now, and an AI answercollapses the whole stage into one result.Where the deal is decidedThe comparison, the proof and the objections achampion has to carry to five other people.Already contested, and already lateBy the time a deal reaches procurement theshortlist was written somewhere upstream.
Top of funnel is now cheap to produce and easy to report. The middle is where a committee decides, and it is the stage almost nobody is funded to build.

Start with what changed on the supply side. Producing top of funnel material used to be the hard part: the blog post, the listicle, the webinar, the infographic. It is now close to free. Any competitor can publish the same volume of introductory content this month, which means volume no longer signals effort, judgment or credibility to anyone. The stage still generates reportable numbers, impressions, sessions, followers, registrations, and those numbers are exactly why it keeps getting funded.

Then look at the demand side. Buyers do not need you to introduce the category any more. They assemble their own understanding from search, AI assistants, peers, communities and review sites, and they do it without identifying themselves. Gartner's research on the B2B buying journey puts the share of the journey spent meeting potential suppliers at roughly 17 percent, with most of the rest going to independent research and to meetings the buying group holds without any vendor present. Gartner, The B2B Buying Journey, checked 4 October 2026.

AI answers accelerate the same shift. Google's own documentation describes how AI features in Search fan a question out into multiple related searches and then compose an answer from the sources they retrieve. Google Search Central, AI features and your website, checked 4 October 2026. The practical effect is not that there is one result instead of ten. It is that the answer gets assembled for the buyer, so the click becomes optional, and the only question that matters is whether your material was among the sources the answer was built from. Being the tenth generic explainer of a well-understood concept is not a path into that set. Being the clearest comparison, the only published pricing model, or the one honest limitation page usually is.

In May 2026 Gartner reported that 69 percent of surveyed B2B buyers turn to sales reps to validate what an AI assistant told them, that 45 percent had used generative AI during a recent purchase, and that buyers consulted an average of seven information sources per purchase decision. The survey covered 645 B2B buyers between August and September 2025. Gartner press release, 20 May 2026. Read that as a brief rather than a trend line: seven sources means the answer has to be consistent across your site, the assistant, the review site and the call, and validation means the human conversation is where the deal is confirmed, not where it is discovered.

Now the part that decides deals. The buying group argues with itself. Gartner's survey of 632 B2B buyers, run between August and September 2024, found that 74 percent of buying teams showed unhealthy conflict during the decision process, with groups ranging from five to 16 people across as many as four functions. Gartner press release, 7 May 2025. Your champion has to carry your case into a room you are not in, against colleagues with different objectives. What they carry is middle of funnel material: a comparison they can defend, a reference story from their own industry, the security answers, the integration list, the business case in their numbers.

So the agency selling impressions is selling the stage that no longer decides anything, priced as though it still does. We would rather build the stage that does: the pages a champion forwards, the proof that answers the objection before it is raised, and coverage of the five to sixteen people who have to agree. Top of funnel still matters, but as a route into the middle, not as the product.

Concretely

What the middle of the funnel is. In deliverables, not adjectives.

Comparison, including against the honest answer

Versus the named alternatives, versus the incumbent, versus building it internally, versus doing nothing. Written so a buyer can tell when you are not the right choice, which is what makes the rest credible.

Proof in the buyer's own shape

A reference story from their industry, at their size, with the same systems. One relevant story beats nine impressive ones from markets they do not recognize.

Objection handling in writing

Security posture, data residency, integration depth, migration effort, support model, what happens if you are acquired. Published, not held back for the sales call.

Buying group coverage

The economic buyer, the technical evaluator, security, procurement, finance and the team that will use it, each reached with the thing they care about rather than one message sprayed at a logo.

The business case a champion can present

The math in their numbers, with the assumptions visible, so an internal advocate can defend it without a vendor in the room.

Material built to be forwarded

One page, no gate, no autoplay, readable on a phone in a meeting. If a champion cannot paste it into a thread, it is not middle of funnel work.

What gets reportedWhat it does not tell youWhat we report instead
Sessions and impressionsWhether any of it reached an account your sales team wantsReach and engagement inside the named account list
Keyword rankingsWhether the ranking page answers a buying question or a curiosity questionRankings and AI citations on comparison, alternatives and pricing-model queries
Marketing qualified leadsWhether sales accepted them, and whether anyone metSales-accepted meetings held, and meeting to opportunity rate
Content published per monthWhether the committee used any of itAssets forwarded internally, and which ones appear in won deals
Account engagement scoresWhether a human ever reached outBuying-group coverage: roles reached per target account

The 80:20 rule, applied

A fifth of the work produces most of the result. The job is naming which fifth, in writing, before the invoice.

The 80:20 rule is easy to quote and hard to apply, because applying it means telling a client what you are not going to do. We do that in the first two weeks of every engagement, channel by channel.

How we cut the scope

The fifth of the work that moves the pipeline number, and the four fifths that can wait the vital fifthDoes the pipeline number move?Comparison and alternatives pages for theterms buyers actually typeOne reference story in the buyer’s ownindustryThe security, pricing model and integrationanswers in writingNamed buying-group contacts in the accountssales already wantsthe trailing four fifthsNobody can say what it changedPublishing on every channel at onceRewriting pages nobody reachesBrand awareness campaigns with no namedaccount listTooling bought before anyone agreed what aqualified meeting isThe test for every line of work: if we stopped doing it this quarter, would the pipeline numbermove? If nobody can say, it belongs on the right.
The same cut in every channel: name the work that moves the pipeline number, write down what is being parked, and revisit it when the number moves.
ChannelThe fifth that moves the numberWhat we park, and say so
AEO and GEOThe twenty to forty prompts a buyer actually asks before a shortlist, answered on pages built to be quoted, with entity and schema basics clean so the page can be retrieved at all.Tracking hundreds of prompts, chasing every engine, and publishing volume aimed at informational questions nobody buys on.
SEOComparison, alternatives, pricing-model and integration pages for your category, plus fixing the handful of templates that already earn attention.Full technical audits of pages nobody reaches, link building at scale, and keyword maps with thousands of rows.
Account-based marketingTier 1 accounts with a named buying group, a researched point of view per account, and one coherent sequence across LinkedIn, email and a human.Display advertising to a five hundred account list, personalization that stops at the company logo, and intent dashboards nobody acts on.
LinkedInTwo or three signals worth reacting to, a founder or executive voice saying something only your team could say, and conversations opened by a person.Connection volume, automated comment pods, and posting every weekday because a calendar said so.
OutboundClean data on the accounts sales already wants, one reason-to-reach-out per account, and a reply handled by someone who can hold a conversation.Volume sending, buying more lists before the first one is worked, and sequences with nine steps that all say the same thing.
Events and webinarsMeetings booked with named accounts before the doors open, and a closed-door roundtable where the buying group talks to each other.Booth traffic, badge scans, and webinar registration counts untied to any account list.
ConversionThe three pages where deals actually start, the demo request flow, and the follow-up in the first hour.Site-wide redesigns, button color tests, and tooling bought before anyone agreed what a qualified meeting is.

The stop test

For every line of work we ask one question: if we stopped this for a quarter, would the pipeline number move? If nobody can answer, it does not get funded this quarter. It gets written down instead.

Depth beats breadth, every time

One channel run properly outperforms four run at a quarter of the attention. We would rather do AEO and one outbound motion well than six channels badly, and we will say so before you sign.

Two channels before five

A second channel only starts when the first one is producing meetings and we can explain why. Adding channels is how programs get busy, and busy is not the same as working.

Parked is written down, not forgotten

Everything we are not doing goes on a visible list with the reason and the condition that would bring it back. That list is as much of the strategy as the roadmap.

Proof

Middle of funnel work, and what it produced. Named clients.

All case studies
“What differentiated Lemniscate Growth is that they started with the buyer. On the inbound side, gated assets built for our category, a maturity assessment tool, and a search strategy that covers SEO alongside answer engine and generative engine visibility. On the outbound side, tiered account lists and multi-touch sequencing. We went from single digit qualified meetings in a month to tens of qualified meetings in a month. They operate like an extension of our own team.”
Sunil Masand
Sunil MasandHead of Product and Marketing, Aavenir

Why top of funnel still gets the budget

Three reasons, none of them about results. It is easy to brief, because volume targets need no judgment. It is easy to report, because impressions and sessions always move. And it is easy to sell, because an agency can show a dashboard going up without ever being accountable for a meeting.

Middle of funnel work has the opposite properties. It requires knowing the objections, which means talking to your sales team and reading lost deals. It produces fewer, slower numbers. And it is uncomfortable to write, because a comparison page that does not admit a weakness is not a comparison page, it is a brochure. That discomfort is the moat: most competitors will not do it.

  • Easy to brief, easy to report, easy to sell, which is why it survives
  • Middle of funnel work needs access to sales, lost deals and real objections
  • The discomfort of publishing honest comparison is exactly why few do it

How the two stages work together

This is not an argument for abandoning the top of the funnel. It is an argument about sequence and proportion. Top of funnel earns the right to be considered: a founder voice that says something non-obvious, a search presence on the questions that precede a purchase, a stage at the event your buyers attend. Middle of funnel converts consideration into a decision a committee can defend.

Run in that order, top of funnel work gets cheaper, because you know which questions matter. Run in reverse, you buy attention and then have nowhere useful to send it, which is the single most common pattern we are hired to fix.

What this looks like in a weekly report

Our reports lead with meetings held, buying-group coverage in target accounts, opportunities created and pipeline value, each attributable to a source. Underneath that sit the leading indicators: citations and rankings on buying-intent queries, replies from named accounts, assets forwarded internally. Impressions and sessions are in the appendix, where they belong.

If a number in the appendix is the only one moving, we say so and change the plan. That is the whole method: a small number of things, measured honestly, changed when they do not work.

Not sure which situation you are in? That is what the audit is for.

The free growth audit maps your funnel stage by stage against your revenue target, then names the two or three moves with the biggest pipeline impact, and the work you can stop.

Questions buyers ask us. Answered plainly.

Still unsure? Ask us directly.

What do you mean by the middle of the funnel?

The stage between a buyer knowing the category exists and a buyer being ready to sign. Concretely: comparison against named alternatives, proof from their own industry, objection handling in writing on security, integration and migration, coverage of the whole buying group, and a business case a champion can present internally. It is the material a prospect forwards to colleagues, which is why we build it first.

Is this just an argument against content marketing?

No. It is an argument about which content. Introductory explainers of well-understood concepts are now free to produce and carry no signal, so publishing more of them changes nothing. Comparison pages, honest limitation pages, pricing-model explanations, migration guides and industry-specific proof are harder to write, get cited in AI answers more often, and get forwarded inside buying groups. We publish the second kind.

How do you decide which fifth of the work to run?

Three inputs, in this order. Your closed-won and closed-lost deals, which tell us the real objections and the real buying path. Your current demand, meaning the pages, queries and accounts already showing interest. And your sales team's named account list. The plan is whatever serves those three fastest, and everything else goes on a written parked list with the condition that would bring it back.

Do you still do top of funnel work?

Yes, as a route into the middle rather than as the product. Search and AI visibility, a founder or executive voice, and the events your buyers attend all earn you the right to be considered. What we will not do is sell volume at the top while the comparison, proof and objection material a committee needs does not exist.

Which use case should I start with if several apply?

Start with the one where the number is most visible to your board, because that is where you will get the room to run the rest. In practice that is usually either traffic without pipeline or deals stalling after the demo. The growth audit sequences them for you, and the first ninety days on any page here assumes you are also fixing measurement, because none of it can be defended otherwise.

How long before middle of funnel work shows up in pipeline?

The honest answer depends on your cycle length, and we will not pretend otherwise. Leading indicators move first: replies from named accounts and citations on buying-intent queries within weeks, assets being forwarded inside deals soon after. Meetings and opportunities follow at the speed of your sales cycle, which in enterprise is usually one to two quarters. We report the leading indicators from week one so nobody is waiting in the dark.

Sources for the numbers on this page (4)

Let’s build your pipeline. Grab 20 minutes with us.

Tell us the revenue number and the market. We will come back with the stages that matter most for you, and the ones you can skip.

  • 20 minutes with a senior operator, not an SDR
  • Bring your revenue target and markets; we bring the pipeline math
  • Slots across US, Canada, India, Singapore and GCC time zones

Prefer email? growth@lemniscategrowth.com

Pick a 20-minute slotStraight to a senior operator. No SDR screen.