The Uncitable Article: Teardown of a Delhivery Blog Draft
Objective: Given a specimen blog-post draft aimed at D2C brands, apply the lesson's GEO content checklist to find every structural defect stopping it from ever being cited by an AI answer engine.
Delhivery's content team published 'How to Reduce Last-Mile Delivery Costs' six weeks ago, aimed at D2C brands researching logistics partners. Manual weekly spot-checks in Perplexity and ChatGPT show zero citations so far, while a competitor's shorter article on the same topic gets cited regularly.
Read the specimen draft's opening, structure, and one core claim, then flag every GEO defect using the lesson's checklist.
What structural defects in this draft are stopping AI answer engines from ever citing it?
Before you start
What you'll need
- —Basic familiarity with GEO content signals (specificity, direct answers, crawler accessibility)
- Interstitial
- an overlay (like a newsletter signup modal) that covers page content, which can block a crawler's headless renderer from reading the article body.
Free path (everything below is enough to finish)
Free, and it's the platform where every citation shows up as a clickable link per the lesson
Free, shareable defect log
The process
Specimens to review
This is the article's opening paragraph. Does it answer the implied question, 'how do you reduce last-mile delivery costs', anywhere in these three sentences?
Logistics has always been the backbone of commerce, connecting sellers to buyers across every corner of the country. Delhivery has spent over a decade building infrastructure that powers this connection, from Gurugram to the smallest pin codes. In this article, we'll explore several dimensions of last-mile delivery cost and share our perspective on the levers available to D2C brands today.
Specimen: synthetic, realistic
Is this claim, which appears mid-article, citation-worthy as written?
Shipping costs are a major expense for D2C brands, and reducing them can meaningfully improve margins.
Specimen: synthetic, realistic
Does this page element affect whether AI crawlers can cite the article?
A full-screen newsletter signup modal appears 2 seconds after the page loads, covering the article body. It has no visible close button on mobile until the user scrolls.
Specimen: synthetic, realistic
Analyze your findings
What to look for
- Answer placement
- Does the opening paragraph answer the article's own title question, or open with company narrative?
- Claim specificity
- Does each factual claim carry a number, source, and year, or just a general statement?
- Crawler accessibility
- Can a headless renderer actually reach the article body, or does an interstitial block it?
- Distractor discipline
- Is a genuinely good detail (an accurate fact, correct grammar, matching audience vocabulary) being mistaken for a defect?
Make the call
The article gets zero AI citations after six weeks while a competitor's shorter piece on the same topic is cited regularly. Given the three specimens reviewed, what is the single most severe defect blocking citation entirely?
Recommendation · Priority: High
“Fix the newsletter interstitial first, since it risks blocking crawler access to the entire article regardless of content quality. Then rewrite the opening paragraph to lead with the direct answer to the title question, and replace the generic shipping-cost claim with a specific, dated, sourced figure. All three fixes are needed, but the interstitial is the one that can zero out citation potential on its own.”
Common mistakes
What trips people up
Flagging the interstitial as only a UX annoyance — for AI citation purposes it's a potential hard blocker, a crawler that can't dismiss it may never read the article body at all.
Missing the buried lead because the opening paragraph is well-written — prose quality doesn't matter if the direct answer to the title question never appears in the passage a retrieval system extracts.
Flagging factually accurate but generic claims as fine — true isn't the bar for citation, verifiable is; a claim needs a number, source, and year to be usable as evidence.
Marking a specific, accurate detail (like a named location) as a defect — distractors are designed to look suspicious; only defects that actually block citation-readiness should be flagged.
Final deliverable
A severity-ranked defect list for the draft plus a rewritten opening paragraph and one rewritten claim.
See a reference example
Zillow blog-draft teardown CRITICAL: Opening paragraph buries the answer to 'how to estimate home value' behind 2 paragraphs of company history -> rewrite: lead with the direct method in sentence 1 MODERATE: 'Home values can fluctuate significantly' (no number, no year) -> rewrite: 'Median home values shifted 4.2% year-over-year in Q2 2025, per Zillow's own home value index' NOT A DEFECT: Author byline with 12 years real-estate experience listed, correctly reinforces E-E-A-T
Success criteria
You're done when you can:
- Correctly identifies all 3 defects with accurate severity
- Does not flag either distractor per item as a defect
- Produces a rewritten opening paragraph that leads with a direct answer
Key takeaway
A citation teardown checks three independent things: can the crawler even reach the content, does the opening answer the implied question, and are the claims specific enough to quote. A draft can fail on any one of these while looking fine on the others, which is why each needs its own check rather than a single overall read.