Andy Ke

Kesha SelectGrowth consultant to the CEO, strategy to release

Daily orders up 41% and monthly revenue up 57% in three months

Store-wide, within three months, not isolated to the cart changes.

As a contract growth consultant reporting to the CEO, I led Kesha Select's cross-functional team across merchandising, inventory and warehouse operations, business development and customer service. I diagnosed where carts stalled, set the strategy and release plan, wrote the specs, and built and tested 19 releases in three months with Claude Code.

Checkout

+30%

Cart-to-checkout conversion, with registration-to-first-order up 38% within three months, associated with the cart changes.

Releases

19 in 3 months

Three cart fixes first, then product pages, navigation, search, promotions and measurement.

Prototypes

10+ in 1 month

Built with an AI-assisted workflow in Claude Code.

How I built it. Built with Claude Code: an AI-assisted workflow helped me ship 10+ prototypes in a month and build each release; I owned the research, specs, release plan, QA and dashboards.

Five checkout hypotheses with verdicts: two confirmed, one partly, two not supported
The five hypotheses I wrote before opening a single recording, the verdict on each, and the source that settled it.
  1. Confirmed: hidden per-campaign minimums and distracting cart recommendations.
  2. Partly: shipping cost, because split orders paid shipping twice.
  3. Not supported: checkout itself (completion was normal) and the in-app browser (no loss at payment).

Strategy and execution, side by side

Each row pairs a call I made with what I shipped, leading the store's teams where the work crossed functions.

Shipping clear before checkout

The call

Shoppers couldn't tell our three shipping methods apart until after ordering, which drove "where is my order" messages and refunds. After studying comparable stores, I decided to show method and timing where shoppers decide.

What I shipped

I led the merchandising and operations teams to organize inventory by shipping method, then shipped color-coded badges on product cards, a delivery timeline above add-to-cart, a cart grouped by method, and a shipping guide page.

Rebuilt menu and categories

The call

I chose to rebuild the menu, categories and subcategories on the product types and tags we already had, so the new structure needed no product data changes and stayed easy to reverse.

What I shipped

I led the business development and inventory teams to ship a new main menu with a brand submenu for partner merchants and new smart collections, all launched behind one setting that rolls back instantly.

Locate the leak

The call

Find where shoppers drop before theorizing. Add-to-cart and checkout completion were healthy; the loss sat between them, in the cart.

What I shipped

Set up Microsoft Clarity with cart segments and filters for repeated frustrated taps, pulled the Shopify funnel by device, and put cart-to-checkout on a daily dashboard.

Hypotheses first

The call

Write the hypotheses before looking at evidence, so the evidence can rule them out instead of confirming a hunch.

What I shipped

Tested all five against session recordings, 15 VIP customer interviews and a customer's cart screenshot. Two were confirmed, one partly, two found no support.

One threshold, explained

The call

Replace several hidden campaign minimums with one sitewide free-shipping threshold that tells shoppers how far they are from it.

What I shipped

Shipped a progress bar that names the gap and a sticky cart, and merged split orders so shoppers stop paying shipping on each one.

Promotions that earn their cost

The call

A discount has to pay for itself. I replaced a game-style unlock with spend-more-save-more tiers, approved on a one-page cost case with break-even modeled first.

What I shipped

Capped the tiers and blocked stacking, so a promotion could not cost more than the case allowed.

Bundles where shoppers decide

The call

Show the deal on the product page, applied automatically, instead of asking shoppers to find and enter codes.

What I shipped

Moved campaigns to automatic bundle discounts and set a coupon naming convention after diagnosing a discount that applied twice.

Cart drawer before and after: three campaign minimums become one free-shipping bar
The cart drawer before and after, recreated in English. Prices are illustrative.
  1. Hidden per-campaign minimums became a single threshold with a bar that names the gap.
  2. Recommendations left the cart; shoppers tapped them and never came back.
  3. Split orders merge, so shoppers stop paying shipping on each one.
Product page with automatic bundle tiers, redrawn; prices illustrative
Automatic bundle tiers on a product page, redrawn. Prices are illustrative, not store data.
  1. Bundle tiers sit next to the add-to-cart button, where shoppers decide.
  2. The discount applies automatically; no code to find or type.
  3. A coupon-name tag shows which promotion applied.

The decision

Chose

Rank every fix by cost and confidence: copy and settings first, layout changes next, developer work last.

Over

A developer sprint on checkout.

Because

Add-to-cart and checkout completion looked healthy; the leak sat in the cart.

Deferred

A Shop Pay test for new customers.

Team

Led merchandising, inventory and warehouse operations, business development and customer service; reported to the CEO.

I owned strategy and execution, from 15 VIP interviews to 19 releases in three months and the dashboards that tracked them.

What retention showed

When the store moved its group-buy regulars onto the new website, only about a third of group-buy-only customers came back. Customers who had ordered on the website once were about twice as likely to stay, so a first website order became a weekly KPI.