<aside> 🛡️
+10% Revenue Per Visitor in 90 days. Or we refund the engagement.
</aside>
Two shopping lists you can buy from today. One for the brand still proving the model, one for the brand defending a $50M run rate. Same categories, different jobs, plus the six questions that decide whether a tool earns its slot.
There are two stacks, and the gap between them is not size for its own sake. It is what each tool has to earn.
At the starter end, every tool either makes you money this month or it goes. At the $50M end, you buy the strongest tool in each category, because a half point of CVR or a point of repeat rate is now worth more than the line item. Your job changes from "buy the cheapest thing that works" to "kill a tool for every tool you add."
Read your own stack against both lists below. Most brands are paying for a category they are too small to use, and missing one they are now big enough to need.
The goal here is complete, not fancy. One tool per job, the cheapest version that does the job properly, and nothing that bills you for scale you do not have yet.
Three rules hold at this stage:
| Function | Pick | Plan / price | Why this one |
|---|---|---|---|
| Store platform | Shopify Basic + free Dawn theme | $39/mo + processing | Deepest app ecosystem, and Dawn is free and fast. Do not buy a premium theme before you have a reason. |
| Page builder | PageFly or GemPages | Free to ~$24-29/mo | Cheap, huge install base. GemPages can clone a layout from a URL when you need a lander fast. |
| Ads | Meta + Google | Free platforms, you pay media | Meta is the center of gravity. Google catches the demand you already created. That is your whole channel mix until both are profitable. |
| Klaviyo Email | From $20/mo at 500 contacts | Bills on active profiles, not sends. Add SMS at the $35/mo Email+SMS tier once your list clears ~1k. | |
| Reviews | Judge.me, or Loox free | Judge.me ~$15/mo; Loox free, paid from ~$10/mo | Judge.me is the budget pick and covers the job. Take Loox free if you want photo-first social proof on day one. |
| Loyalty | Smile.io free | Free, $49/mo Standard later | Only turn it on once you have a real repeat-purchase product. Start free, upgrade when repeat rate justifies it. |
| Subscriptions | Recharge Starter (if consumable) | $0/mo under 50 subs + 1.99% + 19c/txn | Free to start. Add it once a pilot proves customers actually want to subscribe. |
| CX / helpdesk | Gorgias Basic, or Tidio | Gorgias $10/mo (50 tickets); Tidio $39+/mo | Gorgias bills on tickets with unlimited agents, so it scales with the brand, not your headcount. |
| Returns | Loop free Checkout+, or self-serve | Free Checkout+ plan, US/CA labels | A returns app from the start lets you push exchanges over refunds. The free tier covers you early. |
| Analytics | GA4 + Shopify reports | Free | Get these clean before you pay for attribution. They are the baseline everything else gets measured against. |
| Ops / automation | Shopify Flow + Magic | Free, native | Build automations in plain language: auto-tagging, fraud holds, SEO meta. Start here before buying an ops tool. |
Your software floor before media is about $120 to $200/mo. Most of that is Klaviyo and your helpdesk. Everything else is free or close to it.
What you should not buy yet: a paid attribution platform, a forecasting tool, creative analytics, or a dedicated testing platform. A test needs roughly 1,000 conversions a month on the step you are testing to reach significance. Below that you cannot trust most results anyway, so make fewer, bigger swings by eye and keep the budget.
<aside> ⚡
Steal this: open your app billing page tomorrow and find the one tool whose value only shows up at scale you do not have yet. Cancel it this week. On a sub-$1M store, attribution, forecasting, and a paid testing tool are the three most common dollars wasted before they can pay back.
</aside>
Different brief. You have traffic, spend, a team, and a P&L where small percentages are large dollars. Now you buy the strongest tool in each category, you tolerate per-seat and quote-only pricing where the tool earns it, and you accept that the all-in software number is no longer trivial.