The Citation Graph, explained through Lemlist.

Why we ran this study

Lemlist is one of the brands we admire most in B2B. They turned cold email - the most disliked corner of outbound - into something sales teams actually want to talk about.

A 600M+ lead database, $63-$87/user/month, part of the lempire group, holding a 4.7/5 G2 score across 1,396 reviews.

So when we plugged lemlist.com into Wellows and ran 265 tracked prompts across ChatGPT, Gemini, Perplexity, AI Overviews and AI Mode, we expected to confirm what every G2 review already says: category leader, full stop.

What we found was more interesting - and more useful for every founder reading this than a simple "they're winning" headline would have been.

Lemlist is great. The product is great, the brand is great, the customer love is real. They've already done more of the "right" content work than most competitors - 14+ /versus/ pages, a canonical /lemlist-alternatives hub, blog roundups, the works.

And yet, the surface where buyers are now discovering tools - LLM answers - is governed by a different set of rules than Google was. Lemlist sits at a 19.3% Visibility Score, ranked #4 in its category, with 77.8% of AI responses about its space not mentioning it at all.

Volume is no longer the moat. Structure is — and it's the shift most B2B marketing teams haven't yet built measurement for.

We're sharing this study publicly for one reason: the playbook to fix it is the same whether you're Lemlist, a Series A startup, or a one-person brand.

Track citations, not mentions. Restructure the comparison pages you already shipped. Optimize the existing pages before you write anything new. The 90-day plan at the end of this piece works for any of those.

Disclosure: Wellows has no commercial or partnership relationship with Lemlist. This analysis is based entirely on public data and Wellows' own AI Visibility Platform tracking. We chose Lemlist because the brand is widely admired and the gap between its product strength and its AI-citation performance illustrates the thesis cleanly.

The dashboard, in raw numbers

Here's the live snapshot of Lemlist as the Wellows AI Visibility Platform tracks it today:

Lemlist is 708 citations behind Outreach, 170 behind Reply, and 124 ahead of Instantly.ai - a competitor that didn't meaningfully exist three years ago.

The visibility curve from late April shows Lemlist starting near 27% and sliding to ~19% in days. That's not noise - that's a re-ranking event inside an LLM index. Catching it inside a 30-day window is the entire reason monitoring this surface matters.

The per-LLM picture

This is the table that should reshape any marketing roadmap. Per-engine visibility share for the top 5 brands:

Three Things To Notice

1. The Gemini gap is the single biggest delta on the map.

Outreach shows up in 44.4% of relevant Gemini answers. Lemlist shows up in 5.6% - tied with Instantly, ~8x lower than the leader.

Gemini pulls heavily from structured comparison content and authoritative third-party listicles. If Google leans harder into Gemini for commercial intent (they will), this is where the pipeline math gets ugly fastest.

2. Perplexity is more nuanced - and more strategically interesting.

Lemlist sits at 18.8% on Perplexity. On the surface it looks middling, but unpack it:

• It's Lemlist's third-strongest engine, behind AI Overview (21.9%) and AI Mode (20.1%) and ahead of ChatGPT (16.9%).
• Lemlist actually edges out Apollo (18.6%) on Perplexity - the only LLM where Lemlist beats a top-3 competitor head-to-head.
• Lemlist trails Reply.io (23.8%) by 5 points and Outreach (34.4%) by 15.6 points.
• The gap to Instantly is razor-thin: 18.8% vs 18.2% - a 0.6-point lead that could flip with a single review-site update.

Why Perplexity matters more than its raw share suggests: it's the LLM that shows its sources to the user in-line. Every Perplexity citation is effectively a click-through-able mini-listing.

That makes Perplexity citations the most measurable AI-driven traffic source in the stack right now. Outreach's 34.4% there isn't just a vanity metric - it's the equivalent of dominating page-one Google in the ChatGPT era.

3. Lemlist is uniformly mediocre across engines.

There's no engine where Lemlist is a category leader. There's no engine where they're collapsing. They're consistently #4 - which is precisely the position from which a focused 90-day push can produce the biggest absolute gains.

Engine note: AI Overview and AI Mode are tracked separately because they exhibit distinct citation patterns despite both being Google surfaces.

The most important insight: Mentioned ≠ Positively Cited

Here's the realization that should reshape every founder's view of AI Visibility.

Lemlist's sentiment looks fantastic - 651 Positive, 202 Neutral, 15 Negative - roughly 75% positive.

But the mention and the citation are two different things. When you click one level deeper, you see something most tools never show you:

The LLM mentions Lemlist neutrally or positively, while pulling its supporting evidence from third-party pages whose narrative is mixed, lukewarm, or actively praising a competitor.

Lemlist's name is in the answer. Lemlist's story isn't. Two examples make this concrete.

Example A - ChatGPT

"Mentioned: Yes. Sentiment: Neutral. Verdict: Soft."

"is lemlist actually easy to use for a beginner or is that just marketing?"

✓ Brand Mentioned● Sentiment: Neutral

The actual response opens with: "Short answer: it's not just marketing, but it's also not 'effortless beginner-friendly' either. It sits in that annoying middle ground."

It runs a "What users like / Where it breaks down / The reality" structure and lands on "Easy to start, harder to master." The cited sources powering that verdict:

The mention was fine. The base materials weren't. Without surfacing the citation list per response, you'd never know that's where the lukewarm tone is coming from.

This is why "track citations, not just mentions" stops being a slogan the moment you watch a real brand's narrative being assembled in front of you.

To flip this response from "Soft" to "Strong," Lemlist would need to either (a) earn a direct mention in the TechRadar roundup that powers the current answer, or (b) get cited from a third-party source with a more enthusiastic stance on ease-of-use. The first is outreach-heavy; the second is content work — restructuring comparison pages to be guide-like enough that they float to the top of LLM citation patterns. Both are doable in a 60-day cycle.

The 722-opportunity goldmine

Inside the Wellows Outreach Opportunities module, we surfaced a precise number: 722 open opportunities. 0 contacted. 0 won. 0 lost. 0 dismissed.

722 sites already citing Lemlist's competitors - TechRadar, Reddit, G2 lists, comparison roundups - sitting unaddressed.

When we filtered for DA 80-100 alone, it returned 505 high-authority opportunities. The top of the stack:

A few founder-grade observations:

• The DA-100 Reddit threads alone add up to 42+ citations from just the visible top of the stack. Reddit is the single most LLM-cited public corpus on the internet right now - and Perplexity in particular leans on Reddit threads heavier than any other engine.
• The Trustpilot DA-98 entry is Lemlist's own review page — 3 citations on a domain LLMs explicitly trust, on a profile that appears under-leveraged based on review volume and recency relative to G2.
• The TechRadar reviews are the highest-impact single URLs in the dataset — modeled at +13 and +12 citations each if Lemlist is added. A single successful editor outreach to TechRadar represents up to +25 modeled citations — the largest per-touch return in the entire opportunity stack.

Across the 722 opportunities, a conservative average of 2 citations per win and a 20% win rate (~150 wins) yields roughly +200 incremental citations within the 90-day window — enough to clear Reply.io's current 1,165 and approach rank #3. The unbounded ceiling, if win rate or citations-per-win run higher, is materially larger; we model the conservative case throughout this piece.

The 90-Day Plan

The order matters: start with what's already there, then restructure what's already there, then go get what other people own. Each pillar funds the next.

01

Pillar 1 - Days 1-30

Free Citations: optimize what already exists

Estimated impact: +71 to +113 citations - zero new content

This is the lowest-effort, highest-leverage pillar - and the one most marketing teams overlook because it doesn't feel like "doing marketing." You're not writing anything new. You're fixing the citation blockers on pages that already exist.

Wellows' Content Optimization module analyzed lemlist.com:

• 16 pages with optimization suggestions across 7 topics
• Estimated +113 new citations if all suggestions are applied
• That's an 11.4% lift on the entire current 995-citation base - without writing one new article

The single page "2026 Guide to Sales Engagement Platforms" alone is worth +45 citations - a 4.5% jump in total citation count from optimizing one URL.

30-day target: apply suggestions to the top 6 pages - ~+71 citations with zero new content. Call it "free citations" because that's literally what it is - visibility harvested from work already shipped years ago.

Citation lift does not scale linearly across pages — top 6 are weighted toward highest-impact URLs, which is why 6/16 captures ~63% of the modeled total.

02

Pillar 2 - Days 30-60

Content Restructure: audit comparison pages already shipped

Estimated impact: +60 to +80 citations - most counter-intuitive pillar

Lemlist already has a substantial /versus/ directory: 14+ comparison pages, all live, all indexed. By any traditional content-marketing metric, this is best-in-class category coverage.

And yet, Lemlist's comparison content tends to read as self-promotional — opening with declarations of victory and omitting the nuance that LLMs reward.

Shipping the page isn't the work. Earning the citation is.

Why isn't Lemlist's own comparison content getting pulled? Because it reads like it was written for a Google ranking, not for an LLM citation. Every page opens with "Find out why lemlist is the best alternative to X" and concludes that Lemlist wins on every axis examined. LLMs increasingly favor third-party voices over first-party marketing copy when assembling comparison answers — a pattern visible in our own citation data and consistent with how retrieval-augmented systems weight source diversity.

The audit checklist for each existing /versus/ URL:

1. 1
   Open with a neutral verdict, not a declaration of victory. Replace "Why lemlist is the best alternative to X" with "Lemlist vs X: which one fits your team."
2. 2
   Add a "best for / not best for" section that admits where the competitor wins. Counter-intuitively, this makes the page more citable.
3. 3
   Cite named third-party sources (G2 reviews with reviewer names, Capterra, customer quotes attributed to a company) throughout the body - not just in a logo wall.
4. 4
   Replace marketing copy with comparison tables of 8-12 dimensions - neutral language, side-by-side specs, no superlatives. Tables are one of the most-extracted formats in LLM citation patterns.
5. 5
   Add a structured FAQ at the bottom with the actual prompts buyers ask, answered in 2-3 neutral sentences each.

Priority pages to audit, ranked by citation gap cost:

• /versus/lemlist-vs-outreach-io-alternative - Outreach leads at 33.1%. Most expensive page to leave un-optimized.
• /versus/lemlist-vs-apollo-io-alternative - Apollo dominates Gemini at 27.8%. Critical for the engine where Lemlist is weakest.
• /versus/lemlist-vs-reply-io-alternative - Directly addresses competitive comparisons where Lemlist sentiment lags. Quickest sentiment lift in the dataset.
• /versus/lemlist-vs-instantly-ai-alternative - Defensive play; only 0.4 visibility points separate them.
• /lemlist-alternatives - The canonical alternatives hub. Same restructuring rules, higher leverage.

60-day target: restructure the top 6 existing /versus/ pages plus the alternatives hub. A properly restructured comparison page earns 8-15 additional citations within 60 days. That's roughly +60-80 citations from this pillar alone.

Content Creation Opportunities Based on Competitor Insights

Use the Wellows content optimization module to identify what topics your competitors appear in across LLMs, then create optimized, publish-ready content about those same topics. The data below shows the highest-impact content opportunities ranked by estimated citation potential:

Each content piece should be optimized for the specific intent, use cited third-party sources, and target the LLM engines where your competitors are strongest. This approach generates 8-15 additional citations per piece within 30-45 days.

03

Pillar 3 - Days 60-90

Outreach: capture the 722 already-warm pages

Estimated impact: +150 to +200 citations - take what competitors currently own

Now go take the citations competitors currently own. Working from Wellows' Outreach module:

• Filter DA ≥ 80, status = open
• Sort by Est. Citations descending
• Use the built-in "Connect for Mention" workflow - auto-fetched contacts, AI-drafted pitch, customizable

That's roughly +150 to +200 incremental citations over 30 focused outreach days, on top of pillars 1 and 2. Reddit wins double-count on Perplexity because of how heavily that engine indexes Reddit threads. With Wellows including unlimited verified outreach emails on every plan, the bottleneck isn't tooling - it's whether someone on the team owns this 9-5.

The 90-day math, as a founder would write it

Stack the three pillars and the scoreboard looks like this:

Even if execution slips to 50% of model - Lemlist still passes Reply.io on citation count and approaches Apollo on visibility, all on a 90-day clock.

The interesting part of this study? Every number above was already sitting inside the dashboard before we started writing. The plan isn't a strategy invented from scratch. It's the plan the AI Visibility Platform has been quietly assembling since the first prompt was tracked. Someone just has to click "Optimize" on the 16 pages, restructure the 6 existing /versus/ pages, and hit "Connect for Mention" on the 722 warm URLs.

That's the difference between a brand that's mentioned by AI and a brand that's recommended by it.

The bigger lesson for every founder reading this

Lemlist is a great product operating in a surface most marketing teams aren't even tracking yet. The interesting part of the data isn't the 19.3% - it's the visibility swings captured in continuous tracking. Starting near 27% in mid-April and settling at ~19% by early May. Drops of this shape typically correlate with LLM index refreshes rather than day-to-day variance, though confirmation requires a longer time series — which is exactly what continuous tracking provides.

Three rules we're taking from this study into every brand we look at next:

The Lemlist scorecard

Lemlist has the product, the brand, and the customer love to lead this category in the AI era. The data says the work isn't more content - it's better-structured content on the URLs already shipped, and outreach to the 722 third-party pages where LLMs are already looking.

The era of "we'll just rank in Google" is winding down. The next era is "show up in the citation graph." This study was about Lemlist, but the playbook is the same for every brand reading this.

About Wellows