Day 28: Week 4 retrospective - almost at the finish line

Day 28 of 30. End of week 4. Two days left.

#progress #learnings

Day 28 of 30. End of week 4. Two days left.

Let’s take a look of where we are right now.

Interactive week-by-week comparison

See how our metrics evolved over the 4 weeks:

retro - Week 4 retrospective
Day 1
Day 7
Day 14
Day 21
Day 28
2 days remaining
[=] Metrics evolution
Metric Day 1 Day 7 Day 14 Day 21 Day 28
Hours spent 1 14 39 62 78
Lines of code 0 450 2,600 4,100 5,100
Users 0 0 7 38 55
Customers 0 0 0 0 2
MRR $0 $0 $0 $0 $90
[+] What worked
  • Customer conversations (2 calls -> 2 customers)
  • Fast iteration (24h feature turnaround)
  • Performance optimization (+45% faster)
  • Content marketing ("first customer" post)
  • Following up on warm leads
[-] What didn't work (yet)
  • Cold outreach at scale (15 emails -> 0 conversions)
  • Reddit (traffic but no customers)
  • Building features speculatively
[i] Key lessons
Sales is not optional
Every customer came from a conversation, not from landing page traffic.
One customer changes everything
First customer gave confidence, proof, and a feedback loop.
Speed is a differentiator
Both customers mentioned our responsiveness as a deciding factor.
Price is rarely the objection
Decision was about trust and fit, not cost.
Content compounds
Dev log series built audience slowly, then paid off when selling.

The numbers

MetricDay 1Day 7Day 14Day 21Day 28
Hours spent114396278
Lines of code04502,6004,1005,100
Registered users0073855
Paying customers00002
MRR$0$0$0$0$58
Screenshots captured001561,5005,200

We finally got some traction in week 4, and went from $0 to $90 MRR in the last 4 days.

As you can see in our writing, we’ve transitioned quite a bit in our focus. While our initial posts were focused on delivering a working product, our last week’s focus has all been on customer acquisition and promotion.

As you can see in the matrix above, the lines of code in our product hardly changed, and we probably wrote more lines of text in our articles than in our code base in the last few days.

What worked this week

A small reflection on the last week. What worked, what didn’t?

1. Customer conversations. We had two sales calls, resulting in two customers.It would be very optimistic (or unrealistic) to think that we’ll keep having these results, but talking to people works.

2. Fast iteration. We shipped features our customers asked for within 24 hours. That impresses them, and builds trust.

3. Performance work. Making screenshots 45% faster in certain situations made the product feel more professional.

4. Content marketing. The “first customer” post on Indie Hackers drove traffic and signups. We’re not aiming for traffic which converts, though this would be nice: we’re aiming on sharing our experience in building a new product, and hopefully share content people appreciate.

5. Follow-up. The agency deal came from following up on a week-old lead, so it’s important to follow up on these. We’re currently using a Google Sheets document for this, we may need to improve this in the near future.

What didn’t work (yet)

1. Cold outreach at scale. 15 emails, 2 useful responses, 0 conversions. A very low hit rate. Then again, it was low effort work, so perhaps this requires more fine-tuning.

2. Reddit. While we got quite some traffic from Reddit, this didn’t result immediately into more customers. But it’s great to have our name out there.

3. Building features speculatively. Some of the features we built without customer input weren’t what people actually wanted.

Key lessons

Lesson 1: Sales is not optional.

We’re technical founders, but we don’t believe in a product that sells itself, a great amount of marketing is required. So far, every customer came from a conversation, and not from someone stumbling onto our landing page.

Lesson 2: One customer changes everything.

Having our first customer gave us confidence, proof, and a feedback loop. The second sale was easier because we could reference the first success.

Lesson 3: Speed is a differentiator.

Both customers mentioned how fast we responded - to questions, to feature requests, to issues. It’s harder for larger competitors to match this flexibility.

Lesson 4: Price is rarely the objection.

Nobody mentioned that our prices are too high. The decision was about trust and fit, not cost.

Lesson 5: Content compounds.

The dev log series built an audience slowly, then paid off when we had something to sell.

What’s remaining

Features:

  • Core product: ✓ Complete
  • Async mode: Not built (not needed yet)
  • Webhooks: Not built (not requested yet)
  • SDK: Not built (could help conversions)

Marketing:

  • Landing page: ✓ Done
  • Documentation: ✓ Done
  • Content: Ongoing
  • Outreach: Ongoing

Operations:

  • Monitoring: ✓ Done
  • Status page: ✓ Done
  • Support process: Ad-hoc (email)

The goal check

Original goal: First paying customer by day 15. Actual: First paying customer on day 24.

We were 9 days late. But we made the late possible.

Updated goal: 3 paying customers by day 30. Current: 2 customers.

So, we need one more in the next 2 days.

Pipeline

Potential customers who might convert:

  1. Lead from Dev.to - A developer who asked detailed API questions. We sent a follow-up, currently waiting for response.
  2. Free users at 80% quota - They’ve captured 80 screenshots. They might hit limit and upgrade, or at least we can reach out to them.
  3. Agency referral - Customer #2 mentioned they’d refer other agencies. This hasn’t happened yet, but it’s something we can come back to in the upcoming weeks.

If any of these convert, we will hit our goal.

Two days left

The end of our 30 days “in public” track is in sight, and we have only 2 days left:

Our plan:

Day 29: Chase conversions, continue some of our content, and optimize the funnel. Day 30: Our Final retrospective! What’s next?

Book of the day

The Hard Thing About Hard Things by Ben Horowitz

Horowitz was CEO of Opsware through brutally hard times. This book is about the struggles of building companies that business books don’t cover.

Week 4 felt hard. The uncertainty, the slow progress, the question of “is anyone going to buy this?” Horowitz’s message: that feeling is normal. There’s no playbook. You just keep going.

The chapter on “The Struggle” resonated. Building something from nothing is supposed to be hard. If it were easy, everyone would do it.


Day 28 stats

Hours
████████████░░░
80h
</> Code
███████████████
5,100
$ Revenue
█░░░░░░░░░░░░░░
$90
Customers
███░░░░░░░░░░░░
2
Hosting
████░░░░░░░░░░░
$5.5/mo
Achievements:
[✓] Week 4 complete [✓] 2 customers acquired [✓] Pipeline identified
╔════════════════════════════════════════════════════════════╗
E

Erik

Building Allscreenshots. Writes code, takes screenshots, goes diving.

Try allscreenshots

Screenshot API for the modern web. Capture any URL with a simple API call.

Get started