#

Perfectly × HashMatrix
Search & AI Visibility Growth Plan

A two-phase plan for perfectly.so, with delivery-verifiable acceptance criteria for each phase. All diagnostic figures were pulled independently on August 22–23, 2026, without access to your analytics or Search Console.

Client: Perfectly (perfectly.so) Prepared by: HashMatrix Date: August 23, 2026 Version: 1.0
00 / About

About HashMatrix

An AI-native growth marketing agency. We work on four things: search visibility across both Google and AI assistants, influencer and creator distribution, media placement, and reputation management.

Founder

Roger Yin (大罗)

Search & AI visibility specialist

  • 10+ years in search, 200+ projects delivered
  • Eight-figure annual traffic value generated for clients
  • Runs a 100+ site test cluster used to validate ranking factors continuously
Proof

We run this playbook on ourselves

ChatGPT returns HashMatrix as its first recommendation for "fastest growing SaaS SEO agency," and Google's AI Overview cites 大罗 SEO among notable search experts.

Test it yourself, and test the harder phrasings too — head-term coverage is what this plan builds toward and nobody owns all of it. What these show is the mechanism working end to end on a live brand.

01 / Diagnostic

Current state

Pulled independently on August 22–23, 2026, from third-party crawl, backlink and SERP datasets, WHOIS, and live recommendation tests run against ChatGPT and Perplexity. Sources and method are stated at the end of this section.

Strengths

What Perfectly already has (SEO / GEO view)

  • The brand entity resolves correctly. perfectly.so is the top result for both "perfectly so" and "perfectly ai recruiting" in a depersonalized US index. Google already knows which "Perfectly" a buyer means.
  • Four of the nine organic results on the brand query are yours or adjacent — the site, the YC company profile, the LinkedIn company page and a Y Combinator post. That is a base to build a fully owned brand result page on.
  • The YC ecosystem is already an AI citation source. In our Perplexity test, the model cited Y Combinator's recruiting industry directory while assembling its recommendation. Perfectly sits inside a corpus that assistants already read, at no cost. Nothing currently exploits that.
  • Customer proof already exists in public. Named logos (Corgi, LlamaIndex, Mintlify) and a documented displacement story — customers moving off Paraform and cancelling Juicebox. This is the raw material for review profiles and list placement; nothing new needs to be produced.
  • Technical remediation cost is close to zero. Five pages on Framer, no legacy debt, no migration risk.
Open issues

What is missing

  • Absent from AI recommendation answers. We ran the buying question three times — twice on ChatGPT (gpt-5.4, two phrasings), once on Perplexity (sonar-pro), all with web search on. The three answers named Dover, Pin, Ashby, hireEZ, Prometheus Prime, Noxx, Wellfound, Jack & Jill AI, Fetcher, Crucial Recruiting, HyperVelocity, Search Foundry, Systmatic. Perfectly appeared in none of them.
  • What those answers cite is the actual finding. ChatGPT's top pick came from Dover's own blog listicle. Perplexity's twenty citations were almost entirely competitor-published lists, including one on goperfect.com — a brand you already collide with on page one. The models are reading vendor-written lists and repeating them.
  • No third-party review presence. No G2, Capterra or Trustpilot profile found.
  • "perfectly so review" returns nothing about the company. Page one is book reviews, a mattress-topper brand, and a Trustpilot page for an unrelated beauty business.
  • The trust evidence an AI can find is thin. Asked directly whether Perfectly is legitimate, ChatGPT returned a broadly favourable answer — but three of its five citations were ScamAdviser, and the substance came down to "likely safe," a WHOIS privacy note and a young domain. The verdict is fine today. It rests on a scam-checking site rather than on customers, press or reviews, which is what makes it fragile.
  • The domain is about twelve months old. WHOIS shows perfectly.so registered 2025-08-08. Google's trust in a domain accrues with age and third-party citation, and there is no shortcut for the age half.
  • The brand name collides with three competitors. GoPerfect, PerfectRecruit.ai and "Perfect" — the recruiting AI TechCrunch covered raising $23M — all appear on page one for "perfectly ai recruiting."
  • 21 of 24 ranking terms are accidental. Matches on the dictionary word: "perfectly sized," "make it perfectly," "ferpection," "presitely." All 24 point at the homepage; there are no other indexed pages to rank.
  • Brand domain variants sit elsewhere. perfectly.ai, perfectlyai.com, perfectly.io, perfectly.app, perfectly.co and getperfectly.com all resolve to third-party addresses consistent with parking. hireperfectly.com — matching your LinkedIn vanity URL — returns no DNS record.

The numbers

Metricperfectly.soSource & type
Referring domains64Internal Data Source, backlink index — measured
Total backlinks240Internal Data Source — measured
Domain rank (0–100)26Internal Data Source rank. Not Ahrefs DR — the two indices differ and are not interchangeable
Crawled pages5Internal Data Source, crawl — measured
US organic keywords1224 including subdomains; 3 brand-relevant — measured
Est. monthly organic traffic~12 visitsModelled, not measured. Search Console will replace this with the real number on day one
Domain age~12 monthsWHOIS, registered 2025-08-08 — measured
Third-party review profiles0Search across G2 / Capterra / Trustpilot — no listing found
AI recommendation appearance0 of 3 runsChatGPT gpt-5.4 ×2, Perplexity sonar-pro ×1, web search on, Aug 23 2026 — measured, small sample

Method. Everything above was pulled without access to your analytics or Search Console, which means the traffic figure is a model estimate and should be treated as one. Ranking positions come from a depersonalized US index and will differ from what you see in your own browser, where location and history apply — on our call the live checks showed positions 3 and 5 for the brand query against the index's 1. Both are correct measurements of different things. Before anything is contracted, one tool is named and used for both baseline and reporting so this ambiguity does not reach the acceptance criteria.

The AI result is three runs, not a study. It is enough to say Perfectly is not currently being recommended, and enough to identify what the models cite. It is not enough to quantify a share-of-voice number, which is why establishing a proper baseline is the first deliverable rather than a claim in this document.

02 / Opportunity

Market opportunity

Three features of this category decide where the budget should go, and the first one decides how you read every number that follows.

01

Keyword volume no longer sizes this category

AI recruiting is a category in formation, and Google search volume is the wrong instrument for measuring one. The head terms read flat to down while the list-shaped queries climb — "best ai sourcing tools for recruiters" is up 7,100% year over year and hit 720 in July, "best ai recruiting software" up 333%. That is not a category contracting; that is demand moving off the keyword box and into assistants, where no volume figure exists to count it. Our own tests confirm buyers get complete vendor recommendations there without a search ever being logged. What the keyword data still tells us reliably is commercial value, and it is high: $77 to $360 top-of-page bids. Plan against where the demand is going, not against a number that only counts where it used to be.

02

The unit of competition is someone else's list

On "best ai sourcing tool for startups," page one is Gem's blog, Findem's blog, Recruiterflow's blog, Juicebox's own blog, a Reddit thread and three YouTube videos; three vendor homepages appear, none above position three. The AI tests point the same way: the models built their recommendations out of vendor-written listicles. Winning here means being named inside those pages, not out-publishing them.

03

The review layer is unclaimed

No G2, Capterra or Trustpilot profile exists for Perfectly. These carry weight in both rankings and AI answer synthesis, and right now the highest-authority third-party page about the brand is a scam-checking site. One caveat we will resolve before you pay for it: G2 and Capterra are software marketplaces, and Perfectly sells a service. Confirming the category and eligibility is a week-one task, not an assumption.

03 / Audience

Buyer and search journey

Perfectly's buyer is a founder or head of talent at a seed to Series B startup carrying open roles without a full-time recruiter. Search reaches them at two different points, and the two need different content.

Discovery — they do not know you exist

They describe the problem rather than the product: five engineers to hire, no recruiter. They ask an assistant or run a "best tools for" search, and they read the list that comes back. This is where the money is, and where Perfectly currently does not appear.

Content that catches this: third-party listicles, the YC directory, comparison pages, Reddit threads, LinkedIn commentary from people they follow.

Verification — they have met your sales team

A referral, a demo or a LinkedIn post has already happened, and before committing they check whether the brand holds up. Your YC and LinkedIn pages already do much of this work for free, which is why we treat verification as a smaller job than a typical brand-defence engagement would.

Content that catches this: review profiles, comparison-against-incumbent pages, press coverage, an owned answer to "is this company real."

brand + review brand + vs competitor brand + pricing best ai sourcing tools for recruiters ai recruiting agency ai sourcer for startups juicebox alternatives best ai recruiting software embedded recruiter
04 / Benchmark

Competitive benchmark

Juicebox is the closest named comparison from our call. The gap decides which fights are worth picking — and the last row decides the strategy.

Metric (US, Aug 2026, Internal Data Source)perfectly.sojuicebox.aiOthers in the set
Domain rank (0–100)2647Recruiterflow 71 · Findem 70 · Gem 64 · hireEZ 64 · Fetcher 45 · Paraform 40
Referring domains641,340
Total backlinks2407,694
Crawled pages51,046
Ranking organic keywords121,552
Est. monthly organic traffic~12~30,800
Brand search demandno clean brand term isolates*"juicebox ai" 4,400/mo, +406% YoY
Appeared in our 3 AI recommendation runsNoNoDover, Pin, Ashby, hireEZ, Fetcher, Noxx and others did

*"perfectly ai" shows 260/mo but the dataset clusters it under "perfect ai" and flags it as another language, so it is contaminated by other brands and we will not present it as your demand. Clean brand terms are small: "perfectly so" 20/mo, "perfectly recruiting" 20/mo.

The last row is the important one. Juicebox has twenty times your referring domains, two hundred times your indexed pages, and a thousand times your organic traffic — and it was absent from the same three AI answers you were absent from. Domain authority alone is not what gets a brand into an AI recommendation. Being named inside the specific pages those models retrieve is. That is a far cheaper target than out-building Juicebox, and it is why this plan front-loads placement rather than link volume.

On metric sources. These come from our internal data source. Ahrefs indexes the web differently and will show different numbers for the same sites — on Ahrefs, Juicebox reads materially higher than the figures above. Neither is wrong; they are different crawls of the same web. Whichever tool you prefer, we name it in the contract and use it for both the baseline and every monthly report, so the two indices never end up on opposite sides of an acceptance criterion.

05 / Strategy

Core strategy

Two phases. Phase 1 puts Perfectly into the pages that buyers and assistants already read, and closes the trust gaps a buyer hits after a referral. Phase 2 compounds that into owned ranking positions. The order follows the evidence in section 04: placement is what moves AI answers, and it is available now; domain authority takes years and would not have helped Juicebox anyway.

Phase 1 · Months 1–4

Get into the sources buyers and assistants read

Objective: Perfectly is named inside the listicles, directories and review platforms that currently supply both page-one results and AI recommendations — and a buyer who searches the brand name after a referral finds substance rather than a scam-checker.

What gets built

  • Placement inside third-party listicles that already rank page one for the category queries, and that the AI tests showed being cited
  • Full build-out of the Y Combinator directory presence and the other ecosystem pages assistants already retrieve
  • Review platform profiles — eligibility and category confirmed in week 1, then claimed and populated with real customer reviews
  • On-site pages that answer the verification questions: pricing, comparison against named incumbents, customer results
  • Site structure and internal linking, so there is something to index beyond one page
  • Off-site content covering the brand-plus-modifier queries, on domains with existing authority
  • Brand domain variant audit, with the registration list handed to you to action directly

Acceptance criteria

  • Placements live — an agreed number of third-party pages naming Perfectly, each of which ranked page one for a target query at the time of placement. Binary, checkable, contractible.
  • Review profiles live and populated — or, if a platform will not list a services business, a written finding to that effect in week 1 and the budget reallocated before you spend it.
  • Verification coverage — on the agreed brand-plus-modifier query set, the share of top-10 results that are owned or influenced, measured monthly. "Owned or influenced" is enumerated in the contract, not left to our judgment.
  • AI baseline established — the measurement protocol below, running from month 1, with raw outputs delivered to you.
Phase 2 · Months 5–12

Compound placement into owned positions

Objective: convert the citation surface built in Phase 1 into ranking positions and AI appearances Perfectly owns, rather than rents.

What gets built

  • Comparison and alternative content targeting the incumbent names buyers already search
  • Continued placement at higher volume, plus Reddit and community presence
  • Authority link acquisition on a monthly quota, targeted at relevance rather than volume
  • GEO prompt coverage expanded across ChatGPT, Perplexity and AI Overview
  • LinkedIn creator and expert distribution, since your buyers are already there

Acceptance criteria

  • Commercial-intent rankings — keywords in positions 1–20 on a query list agreed in advance, measured against the month-1 baseline in the named tool.
  • Non-brand organic sessions from Search Console, and non-brand demo requests from your own funnel.
  • AI appearance rate on the fixed prompt set, reported as a point estimate with its confidence interval.
  • Relevant referring domains — counted on in-category domains only, not total link count.

How AI visibility gets measured

You told us your acceptance criteria were unclear. This is the part that usually stays vague, so here it is in full, and it becomes a contract appendix rather than a promise in a slide.

  • You write the prompts. The 50-prompt AI set and the brand query set are drafted by Perfectly from your real sales calls and inbound email, before signature. We may propose additions; you hold the veto. A metric we author is a metric we can game.
  • Five runs per prompt, not one. 50 prompts × 5 runs = 250 observations per cycle, which brings the margin of error on an appearance rate near 12% down to roughly ±2 points. At one run per prompt it is ±9 points — wider than any target band worth setting, which makes single-run reporting meaningless.
  • Pinned model versions, named in the appendix, with a stated procedure when a vendor deprecates one mid-engagement.
  • Raw outputs and citation domains logged and delivered to you every cycle, so you can audit the number rather than trust it.
  • Reported with the interval. A move that does not clear the interval is reported as noise, by us, before you have to ask.

What this plan does not claim

  • We do not promise rankings or AI placement. Delivery is contractible and reported monthly. Outcomes are estimates, and the estimate ranges in the pricing table are labelled as such.
  • We will not close the link gap. At any realistic quota, 64 referring domains does not become 1,340 while Juicebox keeps publishing. Links here buy relevance and citation surface, not parity.
  • Brand-query results are hard to attribute. People searching your name mostly arrive via sales, YC and referral. To keep this honest we recommend an attribution field on your demo form that you control and we never touch, and we report non-brand conversions separately.
  • Three AI runs is a small sample. The finding that you are absent is solid; a share-of-voice percentage is not, until the protocol above has run a full cycle.
06 / Schedule

Delivery schedule

Indicative timing for the recommended tier. Exact volumes per month are set at kickoff.

WEEK 1–2
Baseline & feasibility

129-point technical audit. Review platform eligibility answered in writing. Baselines recorded for the brand query set, the AI prompt set, commercial-intent rankings and referring domains, in the named tool. Domain variant list handed over.

MONTH 1–4
Phase 1 · Placement & trust

Listicle placements. YC and ecosystem build-out. Review profiles populated. Verification pages shipped. Site structure and internal linking. First press placement. AI protocol running from cycle one.

MONTH 5–8
Phase 2 · Compounding

Comparison and alternative content. Higher placement volume. Reddit and community. Relevance-targeted links. Expanded prompt coverage. LinkedIn creator distribution.

MONTH 9–12
Owned positions

Commercial-intent rankings mature as authority accrues. AI appearance rate reported against the month-1 baseline with intervals. Non-brand sessions and demo requests become the primary read.

07 / Context

Search is being rewritten by AI

Why the two channels are bought together rather than sequentially.

From searching to asking

Buyers increasingly describe a problem to an assistant instead of typing keywords. "Five engineers to hire and no recruiter" never reaches a results page, and a brand absent from the sources the model retrieves never enters the answer.

There is no page 2 in an AI answer

A results page has ten slots and a second page. The AI answers we ran named three to six companies and stopped. You are in the cited set or you are invisible; there is no partial credit for ranking eleventh.

Winner takes most

Assistants converge on sources they already retrieve. Dover appeared as the top recommendation because Dover publishes the list the model read. Whoever occupies those pages gets cited repeatedly, and each citation reinforces the next.

08 / Insight

What actually gets a brand into an AI answer

The common assumption is that authority carries you into AI recommendations. Our own test says otherwise, and that changes where the budget goes.

Evidence

Authority did not carry Juicebox

1,340 referring domains, 1,046 indexed pages, 30,800 monthly visits — absent from all three runs. If domain strength were the gate, Juicebox would have cleared it.

Evidence

Presence in the retrieved page did

Every company named in those answers appeared because it was written into a page the model fetched: a vendor listicle, a directory, a category roundup. That is the mechanism, and it is purchasable now.

Consequence

Placement first, authority second

Authority still matters — it decides Google rankings and whether publishers and platforms take you seriously. It is the slower half. Buying it first would spend a year before the AI channel moved at all.

One investment, two compounding channels

Paid acquisition stops the day the budget stops. Search and AI visibility accumulate: SEO typically takes six to nine months to take hold and then holds for two to three years, at roughly 10–30% of the cost of buying the equivalent traffic. Placement is the part that can move inside the first quarter; ranking is the part that pays back over years. That six-to-nine-month figure is also why we would rather sell you a twelve-month plan honestly than a three-month one that ends the month before the compounding starts.

09 / Method

A tested methodology

We do not optimize against a generic checklist. We optimize against the factors that measurably move the specific keywords in play.

14,000+
Potential Google ranking factors catalogued
200
High-ROI factors extracted by our own system
TOP 10
Factors that matter most for the current keyword set
100+
Test cluster sites validating factors continuously

129-point site diagnostic

Full technical audit covering indexability, rendering, structure, internal linking and Core Web Vitals.

Real-time ranking factor analysis

We measure which factors are actually deciding the SERP for your keywords right now, not which factors were decisive last year.

One factor system for SEO and GEO

The same analysis drives both the Google work and the AI visibility work, so the two do not pull in different directions.

10 / Process

Deployment process

Five pillars, delivered in sequence and maintained in parallel.

策略
STRATEGY

Strategy

Web quality audit, keyword and topic planning, dual targets for search and AI.

技术地基
FOUNDATION

Technical Foundation

Site structure, internal linking, rendering. The base that serves both Google indexing and AI crawling.

内容
CONTENT

Content & Authority

On-site and off-site content built to rank and to be cited.

权重与品牌提及
MENTIONS

Authority & Mentions

Authority links, Reddit brand seeding, media and press. The off-site signal LLMs check.

AI 可见度
GEO

AI Visibility (GEO)

Prompt and topic coverage across ChatGPT, Perplexity, Gemini and AI Overview, on a visibility dashboard.

11 / Pricing

White Glove Consulting

A fully managed SEO + GEO engagement run by our senior team, delivered monthly and customized to the plan above. Volumes are weighted toward off-site placement, because that is what section 04 showed actually moves AI recommendations, and because a five-page site has limited on-site surface. A line-item reconciliation against our à la carte rates is available on request — we would rather walk you through it than have you reverse-engineer it.

  Foundation SEO GEO + SEO Integrated ★ Full Growth Package
Monthly price $5,000 / mo $8,000 / mo $15,000 / mo
Best for Cold start, foundation only Winning both Google and AI recommendation Full category push
Strategy & technical foundation
Custom SEO / GEO growth strategy
Topical authority planning
129-point technical site auditAt kickoff, then quarterlyAt kickoff, then quarterlyAt kickoff, then monthly
Site structure & rendering fixes
Internal linking build-out
Live reporting dashboard
Placement & trust (Phase 1 core)
Third-party listicle placement on page-1 ranking pages1 / mo2 / mo4 / mo
YC & ecosystem directory build-out
Review platform eligibility finding (week 1)
Review platform setup & seeding (where eligible)
Brand query set tracking25 queries25 queries50 queries
Brand domain variant audit (list handed to you)
Off-site content distribution
Alternative / VS / Review articles (our own, off-site)2 / mo4 / mo8 / mo
Reddit posts + comments6 / mo20 / mo40 / mo
On-site content (site framework support)
On-page optimization of existing pages
On-site content (verification & comparison pages)4 / mo8 / mo14 / mo
Conversion rate optimization (CRO)
GEO / AEO (AI search)
GEO / AEO topic coverage5 topics (50 prompts)10 topics (100 prompts)
Languages covered12
AI platform coverageAI Overview / ChatGPT / PerplexityAll platforms (+ Gemini and others)
AEO visibility dashboard
Brand trust prompt audit (AI assistants)MonthlyWeekly
Off-site authority & brand
Authority backlinks DR30+, in-category4 / mo6 / mo12 / mo
Foundation links20 / mo40 / mo60 / mo
Premium press releaseQuarterlyMonthly
Contracted delivery (checked monthly)
Listicle placements live on page-1 ranking pages, cumulative at month 12122448
AI measurement protocol (5 runs × 50 prompts, raw logs delivered)
Expected results (Estimate · 12 months, not guaranteed)
Keywords in positions 1–20 on the agreed commercial-intent list15–3040–8090–160
Non-brand organic sessions / mo (Search Console)150–350400–9001,000–2,200
AI appearance rate on your prompt set10%–20%15%–30%

Expected results are estimates, not guarantees. Rankings and AI citation rates depend on category competition, product strength, asset quality and Google and LLM algorithm updates — none of which we control. No agency can promise rankings and we do not. The "contracted delivery" rows above are different: those are countable, checkable monthly, and we hold ourselves to them.

Metric definitions. Positions 1–20 are measured on a commercial-intent query list agreed before signature — not total keyword count at any position, which is inflatable with thin pages and tells you nothing. Non-brand sessions come from your Search Console, not from a third-party estimate. AI appearance rate is measured under the protocol in section 05, reported with its confidence interval. One tool is named in the contract and used for both baseline and every monthly report. Different SEO tools index the web differently and will disagree with each other; that disagreement must not be allowed anywhere near an acceptance criterion.

Term — the honest version. The minimum engagement is 3 months, billed monthly, and you can stop there. But this plan runs to twelve months and we would rather say so plainly than let you discover it in month four: at the recommended tier that is $96,000 over twelve months. Placement work should show inside the first quarter, which is what the 3-month exit is there to let you judge. Ranking and AI compounding land later. If after three months the placements have not landed and the AI baseline shows the channel is not reachable for Perfectly, the right decision is to stop, and we will say so.

Recommended tier. GEO + SEO Integrated at $8,000/mo. The $5,000 tier has no AI measurement and no press, which makes it a Google-only engagement — reasonable if you want to test us cheaply, but it excludes the channel section 04 says is winnable soonest. The $15,000 tier suits sites with large page counts and is more than a five-page site needs today.

What we need from you. Search Console and Analytics access, site edit access or a developer contact, a customer list for review seeding, and the prompt and query sets drafted by your team. Budget roughly 3–5 hours a month of founder time for review sign-off; for a four-person team that is a real cost and we would rather size it now.

Scope limits. Casino, cannabis, crypto and adult verticals are excluded from add-on link pricing. Everything published carrying the Perfectly name goes to you for approval before it goes live.

Add-on services

Purchased on top of any tier, or à la carte. Only the modules relevant to Perfectly's plan are listed.

Add-on

Authority link building

Guest post or contextual text link, priced by DR band. Minimum $1,000/month budget. Links here buy relevance and citation surface; at any realistic quota they do not close the gap to a competitor publishing at Juicebox's volume, and we do not present them as if they do.

BandPriceDetail
DR30–DR50$280High relevance, permanent link
DR50–DR60$395High relevance, permanent link
DR60–DR70$650Niche edit on industry-relevant product sites
DR70+$950+Niche edit on well-known product sites, strongest authority effect
  • DR30 minimum — most placements land DR50+
  • Real sites with real traffic — 1,000+ monthly visits, verified in SimilarWeb
  • Editorially placed — each link sits inside relevant published content on the host site, not in a footer, directory or link list
  • Targeted strategy — placements chosen by a senior strategist against your priority keywords, not bought in bulk

Casino, cannabis, crypto and adult verticals are excluded from this pricing.

Add-on

Reddit marketing

Monthly retainer covering posts and comments.

TierMonthlyVolume
Base$1,50025 comments
Growth$3,0004 posts + 50 comments
Scale$5,0004 posts + 100 comments
  • Written by people, disclosed where the subreddit requires it — a human analyst decides what to post and where, and we follow each subreddit's self-promotion rules rather than working around them
  • Established accounts with real history — long-standing community members, not accounts created for your campaign
  • Serves SEO and GEO together — Reddit is heavily cited by both Google and AI assistants, so the same work lifts both
  • US-based Redditors — long-standing community members who know the subreddit rules
Add-on

Press release

Authority media wire distribution, written by senior US journalists with SEO-optimized headlines.

  • Premium — 500+ outlets, $3,000
  • Platinum — 1,250+ outlets, $4,500
Add-on

Media placement

Feature articles in major publications, placed through our contributor network rather than by pitching staff editors. Shorter cycle and higher hit rate than traditional PR.

What it delivers depends on the outlet, and we quote it that way. Some placements are editorial and carry a followed link; others are contributor or sponsored slots that carry a disclosure label and a nofollow link. Both are useful — the second buys brand endorsement and due-diligence cover, not link authority — and we tell you which one you are buying before you buy it. Sample placements from the last 18 months available on request.

Quoted per outlet and per piece.

Add-on

Reputation management

Full-web monitoring, suppression of negative results, crisis response and reputation repair. Covers the scenario discussed on our call: a competitor or a single unhappy customer seeding damaging content against the brand name.

We can suppress negative results, including through legitimate takedown channels handled by US law firms we work with regularly.

Quoted on request, based on the specific situation and channels involved.

Add-on

LinkedIn creator distribution

Your entire buyer base is on LinkedIn, which makes it the highest-leverage single channel for Perfectly. We work with domain experts and creators in the talent and startup operations space to place brand commentary in front of founders and heads of talent, and to feed the off-site brand signal that Google and AI assistants both read.

Quoted per campaign.

12 / Track record

Why HashMatrix

Results delivered across 200+ client engagements.

200+
Client projects delivered
40x
Peak organic traffic growth
+490%
Average keyword footprint growth
4.5x
Average organic traffic multiple

AI-native, not AI-labelled

We run our own tooling for factor analysis, prompt tracking and content operations. The same architecture we discussed on the call: an agent inside the workflow, with human judgment kept in the loop for the decisions that matter.

Four channels under one roof

Search visibility, influencer and creator distribution, media placement, and reputation management. For a brand-defense phase like Phase 1 this matters, because the work crosses all four and coordinating three vendors costs more than it saves.

We report against baselines, not vibes

Every metric in this plan is baselined in week 1 and re-measured on a fixed cadence against the same query and prompt sets. You can check the work without taking our word for it.

13 / Cases

Selected results

Covering both SEO-only and SEO + GEO engagements (some clients buy SEO alone). Shown anonymized or within the scope the client has authorized.

SEO + GEO

Travel OTA

0 → 15%
AI mention share of voice
100x
Search impressions, 10K → 1M
Gemini / AI Overview / ChatGPT · 5 months
SEO

AI SaaS product

18x
Organic traffic
4.2M
Organic clicks, $0 ad spend
4 months
SEO

Beauty e-commerce brand (anonymized)

8x
Organic traffic
5,000+
Ahrefs Total Keywords
200+ keywords on page one · 11 months
SEO + GEO

Home goods e-commerce brand (anonymized)

0 → 18%
AI mention share of voice
5x
Organic traffic
AI Overview / ChatGPT · 9 months
SEO + GEO

Overseas entertainment platform (anonymized)

0 → 20%
AI mention share of voice
10x
Organic traffic
ChatGPT / Perplexity · 7 months
Our own results

Evidence inside AI answers

ChatGPT returns HashMatrix as its first recommendation for "fastest growing SaaS SEO agency." Google's AI Overview cites 大罗 SEO / Daluo SEO as a notable search expert.
Target state for Perfectly: named in the answer to "best AI recruiting agency for startups," where the brand currently does not appear at all.

Case data shown within client authorization and anonymization limits. GEO coverage figures are ranges and shift over time as each LLM platform updates its algorithms.

Why our estimate for Perfectly is more conservative than these cases. These share-of-voice figures come from consumer and high-volume categories, where prompt sets are broad and there are many answers to appear in. Perfectly's prompt set is narrow and the answers we ran named three to six companies rather than ten, so each appearance has to be won against a much shorter list. A 10–20% appearance rate on your prompt set is a harder result than 18% on a consumer travel set, not a weaker one, and we would rather quote what we believe than what matches our best case. Multiples on this page are stated without their absolute bases; ask and we will give you absolutes with dates for any case, plus how long that client stayed.

14 / Clients

200+ brands served

MiniMax
Manus
PromeAI
PixAI.Art
Kling AI
Rabbit
Genspark
Skywork
AiPPT
Talkie
Wanderboat
NeoBrowser
Mureka
Riffusion
FaithTime
SheGlam
Mellow Sleep
PASCAL Design
Spring Financial
PlayNow

Next steps

Book a call and we will walk through the full technical diagnostic and confirm the scope and pricing for your engagement. To start, we need Search Console and Analytics access, site edit access or a developer contact, and a customer list for review seeding.

Book a call