Deepak Gupta

By Deepak GuptaPublished June 23, 2026Content Strategy

Why I Un-Gated All My E-books in 2025 (And What Happened to Pipeline)

MQL-driven gated content is still the B2B SaaS orthodoxy. I un-gated every e-book in 2025. Here are the actual numbers on pipeline, AI citation share, and brand authority 12 months later.

In January 2025 I removed every form, gate, and email wall from my e-book library. Twelve months later, AI citation share is up 4.1x, branded search is up 38%, and pipeline is flat to slightly positive. The MQL count cratered. Nobody who matters noticed.

This post is the actual numbers, the math behind the decision, and a framework for deciding whether to ungate your own library.

The orthodoxy I was rejecting

B2B SaaS content has run on the same logic for fifteen years. You write a long-form asset (e-book, whitepaper, report, benchmark study). You put it behind a form. The form captures a lead. The lead becomes an MQL. The MQL gets scored, nurtured, and routed to sales. Pipeline gets attributed back to the asset. The asset gets reordered up the editorial calendar.

Every part of that chain is now broken, and AI search broke the most important part: the part where the asset is even discoverable.

If your e-book is behind a form, it is not in Google's index. It is not in any LLM training set. It does not get cited by ChatGPT, Claude, or Perplexity. It does not show up when a buyer asks an AI assistant for the best resource on a topic. The asset effectively does not exist outside the small population of people who already trust your brand enough to trade an email.

I covered the structural case for ungating cybersecurity content in why gated whitepapers are killing your AI visibility. This post is the personal experiment that backed up the argument.

The MQL math, honestly

Before I ungated, I ran the actual numbers on what gated assets were costing per useful outcome. The MQL conversation rate (MQL to SQL) on e-book-sourced leads was 4.2%. SQL to opportunity was 31%. Opportunity to closed-won was 19%.

End to end, that is roughly 2.5 closed-won customers per 1,000 e-book downloads. The cost per MQL on paid amplification was running about $42. So one closed-won customer cost $16,800 in MQL acquisition alone, before any sales effort.

The opportunity cost was larger and invisible. Every gated asset was a piece of content that could have been ranked, cited, and indexed. Every form was a friction point that discouraged sharing. Every email I collected was an email that went into a CRM and got nurtured by a sequence that 89% of recipients ignored.

What I did, mechanically

I migrated all sixteen e-books from gated landing pages into a unified reader at guptadeepak.com/ebooks/. Each one became a chaptered, fully-indexed HTML reading experience with proper schema markup, table of contents, downloadable PDF, and a single "share this chapter" link per heading.

I did not add a soft gate, a scroll wall, a "continue reading by signing up" interstitial, or an email-on-PDF-download capture. The PDF is just a PDF. The reader is just a reader. Anyone with a browser can read every word.

I kept a newsletter signup in the footer of each chapter. It collects emails from people who actively want them. That signup converts at a much lower rate than a form gate, and the people who sign up are dramatically more qualified.

The 12-month numbers, three dimensions

Pipeline (the part everyone worries about)

MQL count fell 71% year-over-year. SQL count fell 12%. Opportunity count was flat. Closed-won was up 4%.

The interesting finding: deal size on AI-referred buyers (people who arrived after an AI assistant cited me) was 1.8x the deal size on form-captured MQLs. Sales cycle was 22% shorter. The buyer arrived with context, not curiosity.

The MQL number is the one CMOs panic about. The closed-won number is the one CFOs care about. Mine moved in opposite directions, which is exactly what the AI-search transition predicts.

AI citation share

The e-book content went from cited essentially never (it was behind a form, so models could not retrieve it) to cited 4.1x more often across ChatGPT, Claude, and Perplexity on tracked queries. Specific e-book chapters now show up as primary citations in AI-generated answers to questions like "what is CIAM" and "how should a B2B SaaS founder think about passwordless auth."

The e-books that gained the most citation share were the ones where the gated version had been most popular, which makes intuitive sense. They were the highest-quality assets, and putting them behind a form had been the largest visibility loss.

I cover the mechanics of citation share in the GEO product-content advantage and the cybersecurity AEO playbook.

Brand authority and branded search

Branded organic search (queries containing my name or LoginRadius) rose 38% over twelve months. The mechanism is straightforward: people read a chapter, found it useful, and searched for me later. The branded search rise is the part of the funnel that gated assets actively suppressed, because the asset itself was never the entry point.

The Cybersecurity Entrepreneur e-book in particular drove a measurable lift in cold inbound (founder-to-founder DMs, advisory requests, podcast invitations). None of that pipeline would have existed when the asset was gated.

What the gated approach actually costs in 2026

The honest accounting for a gated e-book in the AI search era includes line items that did not exist three years ago.

You lose the AI citation opportunity entirely. Models cannot cite what they cannot retrieve. Your most authoritative content becomes invisible to the surface that increasingly mediates buyer research.

You lose the indexability lift. Inbound links to a form page do not carry the same authority signal as inbound links to a substantive page. Your domain authority on the topic the e-book covers stays artificially low.

You lose the shareability. People share useful pages on LinkedIn and in Slack. They do not share forms.

You lose the credibility signal. Gating a 40-page PDF on a tactical topic in 2026 reads as either out of touch or as a tell that the content is not actually that good.

You keep one thing: a list of email addresses. That list is worth something, but it is worth much less than it was in 2018, and the cost of generating it has gone up.

The framework: should you ungate?

The honest answer is: probably yes, with three exceptions.

Keep the gate if the asset contains genuinely proprietary data (benchmark surveys you ran, telemetry from your product, customer interview transcripts) and the data is the primary value. Models cannot cite what you do not publish, but proprietary data is what makes you the canonical source on a question. A gated benchmark with 200 responses is more valuable than an ungated rehash of public information.

Keep the gate if the asset is genuinely sales-qualifying (a buyer's checklist that only a buyer with budget and a project would download, like an RFP template). The gate is part of the qualification, not friction.

Keep the gate if you are running a paid-acquisition campaign where MQL volume is the deliverable to the channel partner. Pragmatic, even if it is not a long-term strategy.

For every other e-book in your library, the gate is costing you more than it is worth. Take the assets you currently gate. Estimate the MQL-to-closed-won math the way I did above. Add an honest line for AI citation opportunity cost. Subtract the value of the email list at its current conversion rate. The arithmetic almost always comes out in favor of ungating.

The four reader patterns that emerged

One thing I did not predict: the readership behaviour shifted in ways that gating had hidden.

The first pattern is the chapter scanner. They land on a specific chapter from an AI citation or a Google search, read the chapter, and leave. Gated assets had treated this person as a failed conversion. They are actually a successful read.

The second pattern is the path follower. They start on one e-book (the CIAM buyer's handbook, say), finish it over two or three sessions, then move to a related one (passwordless and passkeys). Internal linking now does work that gating used to actively prevent.

The third pattern is the AI-mediated researcher. They were sent to a specific chapter by an AI assistant in response to a specific question. The conversion intent is high. The session is short. The downstream conversion (newsletter signup, contact form, sales conversation) is unusually likely.

The fourth pattern is the bulk downloader who grabs the PDF and reads offline. Gating had assumed this person was avoiding the funnel. They are actually the most likely to share the asset internally inside their company, which is the highest-leverage outcome of all.

The library, since you asked

The current sixteen ungated e-books are:

Read whichever ones are useful. Share the chapters that help. The MQL gods can wait.

What I would do differently

I would have ungated in 2023 instead of 2025. The MQL framework was already breaking by then, and I left two years of AI citation share on the table by waiting for the data to be unambiguous. The same is probably true for you, right now, today.

I founded LoginRadius in 2013 and we crossed one billion users on the platform. Every cycle in this industry rewards the team that internalized the structural shift one year early. Gated content was the right answer for fifteen years. It is the wrong answer now. The teams that ungate first will own the AI citation surface for the next decade.

Get the newsletter

New writing on identity, AI security, and building software, delivered when it ships. No tracking pixels, no funnels, unsubscribe with one click.