A influencer marketing agency's in-house extension: 1,000 users, 42 five-star reviews, then it stalled

hiCreator is a Chrome extension for looking up influencer profiles on Instagram, TikTok, YouTube, and X, finding emails, and sending outreach. It was listed in August 28, reached 1,000 users by late September, then stayed flat for six weeks. It isn't from an independent developer. It's a service company that received seed investment from Sequoia China and already found influencers for overseas brands. What's worth studying is how a resourceful company cold-starts a new product, and where that cold start runs out. Copying it as a similar extension isn't worth the effort.

October 9: reviews that appeared in one day

On hiCreator's store review page, every visible review was posted on October 9 of 2026: tina yao, Vian Feng, Lee Zhangyu, Zeelot Chen, 张宇, Krystina Lin, Eric Wang, 刘思怡, xiao wang (大鱼)… almost all Chinese or pinyin names. Some read like ad copy: "Small extension, big main-character energy"; some read like jokes: "A colleague recommended this. I'd always wondered how she found so many quality creators so easily... I took her out to a nice meal and she finally revealed her secret weapon"; one review is just two words: "God design."

Third-party extension stats match the curve: the rating count jumped from 4 to 38 around October 11, reaching 42 by October 30, and every one was 5 stars. This basically points to ratings concentrated from the team and acquaintances.

User counts are a separate line: Sep 1 79, Sep 4 302, Sep 15 654, around Sep 25 1,000 (third-party extension stats site). From then until Nov 6 it stayed at 1,000 (Chrome Web Store, 11-01 to 11-05 every day showed 1,000 users / 42 ratings; the store rounds to tiers, so the real figure was between 1,000 and 1,999).

What it is and who made it

What the extension does is clearly described on the store page: in the sidebar, describe the creator you want in one sentence and AI searches for you; filter by country, language, follower count, and views; then "find similar" once you've found a fit; on a creator's profile, see the average and median of recent content, audience country/gender/age breakdown, and fake-follower detection; find their email and send an outreach email directly using a template prepared on the web app; add to a list and export.

But the extension is just the storefront. The main product is the web app hicreator.ai: you submit a brief, and AI runs the whole flow of "find creators → send outreach → follow up → sign contract → review scripts and final cuts → pay → track data."

It's built by the Lessie AI team. The page lessie.ai/hicreator says "From the Lessie team," and Lessie's site navigation lists hiCreator as a "New Product." Lessie is a Deeplink product. Deeplink was founded in August 2022 as an overseas influencer marketing service provider. Its homepage lists clients including CapCut, Skywork, PixVerse, Monica, PLAUD, UGREEN, ECOVACS, and DREO. It has raised angel, Pre-A, and Pre-A+ rounds, with investors including Tec-Do, Sequoia China Seed Fund, Linear Capital, and K2VC (deeplink.media/about).

In other words, hiCreator is a service company that runs paid campaigns for brands, and it turned its internal tools and creator database (claimed 50M+) into SaaS.

Where the money comes from

The extension itself doesn't make money. The revenue comes from the web app, sold as RMB credit packs (hicreator.ai/pricing, 11-06):

Credits don't expire, and invoices are available: WeChat and Alipay payments can get a domestic invoice, and Stripe works too. Single tools are charged by usage: AI deep search costs 1 credits per creator, email lookup 0.2 credits per link, audience analysis 40 credits per run, fake-follower detection 30 credits per run, and each workspace gets 100 free outreach emails per day. Fully managed service costs 25 credits per creator, roughly ¥1.9 each at the Growth price (estimate).

The pricing page also cites its own conversion numbers: about 20% reply, 8% quote, 4% sign. At those rates, the Growth plan runs 2,600 outreach attempts and signs about 104 collaborations, so the platform fee works out to roughly ¥48 per signed deal (estimate, based on its own reported conversion rates, excluding payments to creators).

A few details show who it sells to: RMB pricing, invoices, credits that don't expire, and tiers by team seats. This is built for domestic teams going overseas that need expense reimbursement and procurement approval. One thing to note: it talks about "charging by results," but email lookup is charged per valid link, "even if no public email is found" (original pricing page text), so it isn't strictly "pay only when found." How many paying Web customers it has and how much it has collected isn't visible from outside.

How the 1,000 users came in

All from the company's own resources; none of it came organically:

There have been no posts about it on X in the last 90 days, it doesn't show up on TikTok search, and it's absent from Product Hunt and Hacker News.

The timeline is clear: the first wave came from the launch push in the first week of September, the blog, Qiita, and directory listings spread through the middle and late part of September, and the October reviews and paid reviews were aimed at making the store page convert better. After 1,000 it went quiet for six weeks, which suggests the self-owned channels produced one-off traffic that ran out. Reviews can push the rating up to 5.0, but they can't carry the user count.

Who it's competing with

This kind of extension has already been done to death:

(User count from the Chrome Web Store, 11-06.)

Profile-data overlays, email finding, lookalike search, fake-follower checks, and exports are all offered by the players above, mostly for free. hiCreator's real differentiation is only on the web side: AI that runs the full flow from outreach to payment, RMB-invoiceable credit packs, and conversion data from its service company. But "AI agent runs the whole workflow" is also what AhaCreator and Nox are saying.

If it were you

C: Learn their cold-start playbook, but don't build the same kind of extension. An indie developer building a creator-data extension would be going head-to-head with a crowd of similar products that already have 10k to 100k users and are mostly free. The things that actually have value for it, namely a 50M creator database, a service company's client and conversion data, and B2B sales that can be invoiced, are all out of your reach. Even with those resources, the extension has stalled at 1,000.

Who it's for: Small teams already running a service in some industry (paid ads management, outsourced operations, export-trade fulfillment, recruiting headhunting) with their own clients and internal processes, who want to turn a tool they use every day into a product.

Who it's not for: Indie developers without a creator-data source or B2B clients, and anyone hoping to build traction from organic Chrome Web Store traffic. The store is already crowded in this category.

How to do it: Copy its structure, not its category. Turn the most repetitive part of your service (finding people, sending outreach, following up) into a tool, and first give it to your own clients. For pricing, charge per result or deduct credits per usage, and offer one-time credit packs that don't expire and can be invoiced, which lowers the purchasing barrier for domestic teams. If you still want to touch creator marketing, go after a small slice that both it and its rivals cover poorly, like one country, one language, or one platform (its audience analysis and fake-follower checks don't support X), rather than building yet another all-platform creator database.

How to acquire users: The cheapest move is one its sister site is already using: publish a batch of free, login-free micro-tools (fake-follower checks, engagement calculators, email finders) to capture search traffic, and place an entry point to your main product on those pages. Fake reviews and paid reviews from small accounts might make the store page look better, but don't expect them to bring in users.

Is there still a window? No, not really. This category has had players for 2016 years, and the AI agent layer has been talked about for 2025–2026 years, so first-mover advantage is gone.

What would show I'm wrong: ① hiCreator's user count crosses the 2027 tier before the start of 5,000 and real reviews from non-acquaintances show up, which would mean the web side is pulling the extension along and the product itself has taken off; ② it or a deep link publicly discloses the number of paying web-side clients or revenue, and the figures are substantial, which would mean the "service company builds SaaS" path works in this niche and is worth revisiting; ③ conversely, if it stays at the 1,000 tier through year-end and extension updates slow down, that confirms it was a resource-driven cold start with nothing behind it.

One line for everyone building products: company resources can push a new product to the first 1,000, but they can't push it past that. Growth beyond 1,000 comes from the product itself.

Data and sources

• Store data: hiCreator's detail and review pages on the Chrome Web Store (11-06:1,000 users / 5.0 stars / 42 ratings, v0.2.46, developer marked as non-trader, with in-app purchases, 5 languages); daily snapshots from 11-01 to 11-05.
• User and rating curves: third-party extension statistics sites (9-01 to 11-06); the store rounds to tiers.
• Pricing: hicreator.ai/pricing (11-06), including the original conversion rate and billing rules; not verified through a purchase.
• Company background: lessie.ai/hicreator, lessie.ai homepage navigation, deeplink.media/about.
• Channels: LinkedIn company page and employee posts, Qiita (2026-09-01), lessie.ai blog (2026-09-18), YouTube reviews from Invente com IA and Prof. Ken, PeerPush / AppRater / SourceForge listing pages; searches on X for the last 90 days, TikTok, Product Hunt, and Hacker News returned nothing.
• Competitors: individual extension pages on the Chrome Store (11-06); the scale of AhaCreator and Nox comes from media coverage and their own websites.
• The claims that deep-linked existing clients moved over and the per-signing cost are inferences or estimates; web-side paying clients and revenue have not been disclosed.