# Ronak Shah

Ronak Shah shares practical lessons on growing direct-to-consumer brands, including ways to use AI for creator sourcing, customer retention, and other growth tasks. His posts draw on experience scaling Obvi past $100 million in revenue.

- Profile: https://b2binfluencer.marketing/creators/ronak-shah/
- LinkedIn: https://www.linkedin.com/in/ronak06
- Median of last 7 posts: 51 reactions, 18 comments (checked 2026-11-02)
- Topics: artificial-intelligence, ecommerce, creator-marketing, ai-in-ecommerce
- Based on 8 dated posts, latest 2026-09-24

## Summary

Most useful for practical advisory on AI-assisted creator sourcing and repeat-purchase retention for DTC brands.

- Best for: It fits a DTC brand past $10M seeking creator acquisition and repeat-order analysis.
- Not for: This is not for paid creator posts, short video reviews, or reach buys.
- Check first: Confirm data access, deliverables, and how success will be measured.

### Takeaways

- A rising revenue line can hide a customer base that never comes back.
- Subscriptions manage only about a fifth of repeat customers.
- Creator programs stall at twenty people because sourcing does not scale by scrolling.

## Main ideas

- **The prompt is the lesson: teaching AI use by handing over the exact instruction**: In a recurring series ("one way I'm using AI every day"), each post takes a single operational problem and publishes the prompt used, so a reader can rerun it on their own inputs. The argument built into the format is that the tool is not the interesting part; the named inputs and the required output are. Ronak Shah's posts specify what goes into the model and what should come back — full order history, a day's video transcripts, a product and contact target — instead of describing AI in general terms.
- **A rising revenue line can hide a customer base that never comes back**: The argument is that top-line growth and a leaking customer base look identical until you pull the cohorts, so the first move on a retention problem is a repeat-purchase curve rather than a loyalty program. New customers entering at the front door can mask the ones walking out the back, and a revenue dashboard will show that as growth.
- **Subscriptions manage only about a fifth of repeat customers**: Shah argues that even the best subscription programs convert roughly 20% of repeat customers, leaving 80% who reorder on their own with no cadence and no one managing the relationship. Treating subscriptions as the whole retention story therefore leaves most repeat revenue unmanaged. The direction Shah points to is predicted rather than scheduled commerce: using data to judge when a specific customer is ready to reorder instead of nudging everyone on the same cadence.
- **Screen a decision before the money goes out, not after a live test fails**: Two posts make the same case for expensive, hard-to-reverse decisions: feed the model your own winners and losers and ask which bucket a new idea resembles before spend goes behind it. For financial models, Shah separates safe errors from dangerous ones — a wrong assumption can be defended, but an internal contradiction across tabs tells a sharp investor the founder does not know their own numbers, and it is exactly the kind of error that stops being visible after the hundredth look.
- **Creator programs stall at twenty people because sourcing does not scale by scrolling**: Shah argues the difficult part of a creator program is finding hundreds of small accounts that genuinely fit the product, which cannot be done by scrolling a feed — and that this is why most programs stall at around 20 people. The method is to point AI at a category, pull nano and micro creators who already post about similar products, and score them on fit, real engagement, and whether they drive sales rather than on follower count.
- **In regulated categories the bottleneck is clearing creator content, not making it**: Shah argues creator programs are treated as a volume game, but every video makes claims, and in a regulated category one unsubstantiated line is a legal problem waiting to happen. The practical cap on how much creator content can safely go live is review capacity, so that is where the automation belongs — not in producing more content.
- **Try a partner's product on your own brand first, and treat the criticism as the proof**: Shah's stated standard for backing a company publicly is to run their own brand through it first, then report the unflattering parts. The reasoning: a sales tool would flatter, so an analysis that scores your moat "Basic" and names where you are exposed is the reason to trust it. The same post argues about incentive design — a partner who funds the test and earns only a percentage of sales shares the downside, which Shah calls unusual in more than a decade of DTC.
- **Partnerships are judged on remembered details, not the pitch**: Shah's conclusion from a partner visit is that what actually counts is paying attention to what someone cares about and then showing up — not the deck and not the follow-up email. He applies that standard to how a partner treats a relationship, and by extension how he would judge one.

## Work to read

- [Finding 200 nano and micro creators without scrolling TikTok](https://www.linkedin.com/posts/ronak06_i-committed-to-posting-one-way-im-using-activity-7508925807789158404-OStc)
- [The repeat-purchase curve that flipped a growing brand's story](https://www.linkedin.com/posts/ronak06_i-committed-to-posting-one-way-im-using-activity-7516135868252971008-aw2y)
- [The 20% subscription ceiling and the customers nobody manages](https://www.linkedin.com/posts/ronak06_heres-a-number-that-bothers-me-the-best-activity-7508902545751465984-CgP-)
- [The ugliest landing page won by $228 a customer](https://www.linkedin.com/posts/ronak06_i-committed-to-posting-one-way-im-using-activity-7516543459323998209-P0t-)
- [Clearing 40 creator videos for compliance in minutes](https://www.linkedin.com/posts/ronak06_i-committed-to-posting-one-way-im-using-activity-7517946427013115905-bgmV)
- [Ten contract manufacturers in one afternoon](https://www.linkedin.com/posts/ronak06_i-committed-to-posting-one-way-im-using-activity-7520855595718000641-vaSR)
- [The market audit that scored his own brand "Basic"](https://www.linkedin.com/posts/ronak06_zygpartner-activity-7515765577886404611-J4OF)
- [Three errors an investor would catch in the first meeting](https://www.linkedin.com/posts/ronak06_i-committed-to-posting-one-way-im-using-activity-7521933963083321344-B3_h)
- [A dinner, a candy box, and how a partnership got tested](https://www.linkedin.com/posts/ronak06_nikolaus-from-whop-flew-from-california-to-activity-7521568949583630337-5sVF)

## Brand relationships

- **Obvi** (owned_business)
- **ZyG** (disclosed_partnership)
- **Bold Commerce (rePete)** (disclosed_partnership)
- **Whop** (disclosed_partnership)

## Partnerships

Ronak Shah has not published a sponsorship policy, media kit, or rate card, so there is no documented package a brand can simply buy. The posts show two kinds of commercial collaboration. Ronak promoted ZyG, a market-research platform that launched out of stealth with $118 million raised, in a founder-voice post carrying the #zygpartner hashtag and a link in the first comment; before posting, Ronak ran a hero Obvi product through the platform's short intake and reported the results that came back days later, including an unflattering score that labelled Obvi's moat "Basic." Ronak also co-produced a free gated guide with rePete by Bold Commerce on repeat purchasing beyond subscriptions, using figures such as Built Brands growing reorder revenue four times faster than subscription revenue in 60 days. Both examples are teaching-first and tied to a named product or service; neither discloses terms, deliverables, exclusivity, or compensation, and a partner hashtag alone does not establish payment. One post about a dinner with Whop shows how partnership conversations reach Ronak — relationship-led, with the partner acting on a small personal detail within two days — rather than through a published intake form. Treat any format as something to negotiate directly, and expect to give Ronak enough access to a real product, account, or data set that he can say something specific rather than generic.

## Alternatives

- Jonathan McGraw: Choose Jonathan McGraw when the brand's real blocker is proving that creator and TikTok Shop spend pays off across the whole business, not building the creator list itself. The published work is cross-channel measurement: portfolio data from Prime Day where product-detail-page views fell while conversion, orders and ROAS improved, an Adobe figure that AI-sourced traffic to US retail sites rose 89% year on year and converted 40% better, and a SmartScout finding that the top 10,000 Amazon brands with a TikTok Shop presence grew Amazon revenue 5.2 times faster than those without one. The framing question is how much of the purchase decision was won before the shopper reached the product page, which is the question a finance lead or board asks when a channel gets renewed or cut. Use this when the deliverable is a defensible cross-channel read on whether creator investment is working.
- Jake Bjorseth: Choose Jake Bjorseth when the brand is already committed to TikTok Shop and the problem sits inside the channel: ad dollars re-buying customers the brand already owns, and the limits of the platform's main ad product at higher spend. The published argument is that the in-platform repeat rate should be driven low for net-new acquisition, with one named brand climbing from a 30% repeat rate in January to 72% in August, meaning most ad spend stopped acquiring new customers and started re-buying ones who would have bought elsewhere. The companion breakdown of GMV Max lists the constraints plainly: no retargeting, no objective other than conversion, no way to exclude existing customers, and no way to run specific creatives at specific targets. It then gives concrete mitigations: differentiate assortment between channels and price TikTok Shop higher, cut 5-10% of ad budget to contract creators around new SKUs and run reach ads to bring net-new consumers into the algorithm, and retire creatives before fatigue sets in. Use this when the next 90 days are about making TikTok Shop itself more profitable and less cannibalistic.
- Samy Waiche: Choose Samy Waiche when the brand has already decided retention is the constraint and wants a membership or subscription product built and run, especially in a category where repeat delivery of the same item does not fit how customers shop. The published case is the Adore Me model: members paid $39.95 a month, received $39.95 in store credit, and chose what to buy and when, which was tested across millions of transactions before becoming Subscribfy, an all-in-one retention platform. The public position is that improving retention changes the economics of growth, and that paid membership with store credit, member pricing and shipping benefits gives customers a reason to keep choosing the brand beyond the first purchase. Use this when the deliverable is a working membership programme with defined member benefits, not an analysis the team runs itself.

---
AI-assisted summary of public posts. Audience demographics are not available. Engagement appears only where we checked recent posts. Method: https://b2binfluencer.marketing/methodology/