What I'd Do Differently If I Started Cozy Sips Again (3 Years In, Honest List)
Three years into running Cozy Sips, I have a list of things I'd do differently if I started over. Not the inspirational "here's what I learned" version — the operational, financial, and personal list. The mistakes I made that cost money, time, or sleep. The decisions I'd reverse. The things I'd start sooner.
This is the follow-up to our what I'd tell someone who wants to start a coffee company post. That post was advice for someone starting. This one is the lessons I learned by doing it wrong the first time.
1. I would not launch with a 15-SKU catalog
We launched with 15 SKUs — 8 coffees and 7 teas. The thinking was that variety gives customers more reasons to buy. The reality was that 15 SKUs meant 15 inventory lines, 15 packaging SKUs, 15 product descriptions, 15 SEO pages, and 15 things to maintain, rotate, and account for.
In year one, we discovered that 5 SKUs accounted for 70% of sales. The other 10 were either niche items that sold slowly or items I thought customers wanted but didn't.
If I started over, I'd launch with 5 SKUs. The 5 that I knew sold. The 5 that I had tested with a small customer base. The 5 that I could source reliably and consistently.
I would add SKUs only after the original 5 had proven repeat purchase and after I had the operational capacity to handle more. Adding SKUs before you have operational capacity is a fast way to create inventory chaos.
2. I would set up the subscription model in year one, not year three
We don't have subscriptions. We've talked about it. We've planned it. We've decided three separate times that "now is not the moment." Now, three years in, I'm setting it up.
If I started over, I'd build the subscription infrastructure (recurring billing, customer portal, churn management) in month one. Subscriptions are the foundation of a sustainable specialty coffee business. They smooth cash flow, build customer lifetime value, and create the kind of predictable revenue that lets you take risks.
I spent three years trying to build subscriptions as an add-on to a one-time-purchase model. That was the wrong order. The right order is to build subscriptions first and treat one-time purchases as a complement.
3. I would not underprice in year one
Our year-one prices were too low. I priced based on what I thought customers would pay rather than what the coffee actually cost to source, roast, package, and ship.
The math I missed: a $19.99 bag of coffee has roughly $7 in green coffee cost, $2 in roasting, $3 in packaging, $4 in shipping, and $2 in platform fees and payment processing. That's $18 in costs on a $19.99 sale. The remaining $2 is gross profit, and after customer service, returns, and overhead, the net is closer to zero.
We raised prices in year two. Sales didn't drop. The lesson: customers who buy specialty coffee are not as price-sensitive as I assumed. They want quality, reliability, and a story. The right price is the one that keeps the business sustainable.
If I started over, I'd price based on the cost-plus-margin formula from day one, not based on competitor benchmarks or fear of customer pushback.
4. I would build the email list before launching the website
We launched the website and then started building the email list. The list grew slowly for the first year because we were capturing only the customers who found us organically.
If I started over, I'd build the email list first — through content, partnerships, giveaways, and lead magnets — and then launch the website with a list of 500+ interested people ready to buy on day one.
The math: a website launch with 500 engaged email subscribers typically generates $5,000-$15,000 in first-week sales. A website launch with 0 engaged email subscribers generates whatever organic traffic happens to land. The first launch is much more forgiving than the second.
5. I would not do my own customer service for the first 6 months
This one is counterintuitive. The advice I got was "respond to every customer yourself for the first year to learn what they actually want." I did that.
The result: I learned what customers want, but I also learned that customer service is the single biggest time sink for a solo founder. I spent 15-20 hours per week on customer service in year one. That's 15-20 hours I didn't spend on sourcing, product development, or growth.
If I started over, I'd outsource customer service in month one (even at a higher cost per hour) and learn from transcripts and patterns rather than from direct conversations. The lesson is still valuable — but the time cost is too high for a solo founder who needs to be working on the business, not in it.
6. I would set up Google Search Console in week one, not year two
We didn't have Google Search Console set up until year two. That meant we didn't know what people were searching for, which of our pages were ranking, which queries were generating impressions, or where the technical SEO issues were.
If I started over, GSC would be set up before the first blog post was published. The free data from GSC is the single most valuable SEO tool for a small business, and not having it means flying blind.
7. I would not start with a Shopify theme that looked premium
We launched with a premium Shopify theme. It looked great. It cost $350. It had 200+ features we didn't use.
The reality: the premium theme added complexity, slowed down the site, and made every customization harder than it needed to be. We eventually switched to a simpler theme and the site got faster, the customization got easier, and the conversion rate went up.
If I started over, I'd start with the simplest, fastest Shopify theme I could find. Customization and visual polish can come later, after the operations are stable.
8. I would not try to be on every social platform
We launched with presence on Instagram, Facebook, Twitter/X, TikTok, Pinterest, Reddit, and YouTube. The thinking was "be everywhere." The reality was that each platform requires different content, different cadence, and different engagement patterns. We couldn't maintain quality on all of them.
If I started over, I'd pick 2 platforms and do them well. Probably Instagram and one of TikTok or Pinterest. The other platforms would come later, after the first 2 were generating consistent engagement.
9. I would not raise prices in increments
When we finally raised prices in year two, we did it in small increments ($1 at a time) over 6 months. That was the wrong approach. The constant price changes created customer confusion, customer service overhead, and platform fee variability.
If I started over, I'd set the right price from day one and raise it in larger, less frequent steps ($3-$5 at a time, no more than twice a year). Each price increase would be communicated clearly with a reason, and customers would adapt faster than I expected.
10. I would start the rescue partner support earlier
We didn't formalize our rescue partner support until year two. In year one, we said we supported rescue partners but didn't have a formal program.
If I started over, I'd set up the rescue partner program in month one — even if the monthly contribution was small. The structure (who, how much, what % of profits, how the partner is selected) needs to be in place before the sales start. Adding it later creates operational and accounting complexity that's avoidable.
11. I would not skip the personal story in marketing
Our year-one marketing was very product-focused. We talked about the coffee, the origins, the brewing. We didn't talk about me, the founder.
The result: customers didn't connect with the brand. They liked the coffee, but they had no reason to choose us over the next specialty roaster. In year two, I started writing more personal posts (the "founder voice" posts, the B-t-S posts, the lessons-learned posts), and customer engagement went up measurably.
If I started over, I'd lead with the founder story from day one. Customers want to know who's behind the coffee. They want to know why the founder cares. They want to feel like they're supporting a person, not a brand.
12. I would not buy inventory in bulk to "save money"
We bought 6 months of green coffee inventory upfront because the supplier offered a volume discount. The discount was 8%. The cost of holding that inventory (storage, capital tied up, cash flow impact) was much higher than 8%.
If I started over, I'd buy inventory in smaller batches more frequently, even if the per-pound cost was slightly higher. The flexibility and cash flow benefit is worth more than the volume discount.
The takeaway
The pattern across these 12 lessons is consistent: I made the mistakes that founders make when they're optimizing for the short term (low prices, broad catalog, multiple platforms, bulk inventory) instead of for the long term (sustainable margins, focused catalog, fewer platforms, just-in-time inventory).
The hardest lesson was number 11 — that customers want a story, not just a product. The most expensive lesson was number 12 — that bulk inventory discounts are often more expensive than they look. The most actionable lesson was number 6 — that GSC should be set up in week one.
If you're thinking about starting a coffee company, the what I'd tell someone who wants to start a coffee company post covers the advice I give to others. This post is what I'd tell myself.
Brew slow. Sip slow. Learn out loud.
Why we don't run sales (a related pricing philosophy): /blogs/news/why-we-dont-run-sales
Why we don't sell decaf (a related product decision): /blogs/news/why-we-dont-sell-decaf