logo New Idea

April 2022 Retro

I used Canva once. Now I'm a graphic designer.

In last month’s retro , I felt sorry for myself and for my performance as an entrepreneur. It was a low point for me, but I’m happy to move past that.

Since then, I’ve tried my hand at some not-coding. Specifically, graphic design with the purpose of demonstrating product features that don’t exist yet.

Up until now, it just seemed easier to code the damn feature instead of trying to figure out how to use Photoshop Photopea . But there is a third option that I haven’t tried before: using an entry-level design tool to make quick & dirty concepts.

I took full advantage of my noob status as a graphic designer to joyfully commit design atrocities, as I recreated AdSlicer’s landing page.

Last month’s goals

  1. Spend more time on marketing & selling AdSlicer than coding.

    Final grade: 👍

Main events

Coding

I silently deployed an Ad Performance over Time graph and it’s already my favourite graph (what’s your favourite graph? 😄).

A visualization of conversion rates over time for one of the Facebook Ad accounts that my partner manages. A visualization of conversion rates over time for one of the Facebook Ad accounts that my partner manages.

The cool thing about this graph is that it tells me if my ad metrics are improving or getting worse. This gives me the opportunity to course correct.

I built this graph because a) I was curious, and b) it was the foundation of (what I thought would be) AdSlicer’s next killer feature: Keeping track of ad optimizations and visualizing their effects.

Not-coding

From my previous customer interviews, I’ve determined that AdSlicer’s angle of “visualize your ad results” wasn’t compelling enough and wasn’t solving a problem people really cared about.

I needed to redo my landing page to resonate with my customers (i.e. Facebook Ad marketers) by describing their pain in their own words.

The problem was that I’ve already built something to solve “visualizing ad results”, but now I need to describe & demonstrate a new solution that doesn’t exist yet.

Enter not-coding. It’s like regular coding, but less fun and less fulfilling and it doesn’t do anything.

I needed this not-coding restriction to avoid falling into the trap of building something people don’t want. And I’m thankful that I did.

Graphic design

I started with the landing page copy by extracting words + phrases from my interview notes to craft AdSlicer’s new angle of “Take the guesswork out of Facebook Ads”.

With the copy done, I needed to visually represent features that would achieve the outcome. Again, these features don’t exist, so wtf do I do? I lie!

I used Canva to (visually) build things that don’t exist yet. And it was amazing to do!

It was like writing my first program. I could think of anything and bring it into a plane of existence 🤯.

You can find the rest of my proudly terrible designs on the newly released AdSlicer website 👇

Market research

At this point, it feels like I’ve done everything in reverse order. As part of my not-coding work I decided to really really really explore the market landscape and see what other tools are out there.

Each one of my customer interviews (3) reported that they don’t use any tools to help them with Facebook ads. Useless and not a good sign.

The go-to place for Facebook Ad marketers to hangout are in various Facebook groups. The equivalent subreddits seem quiet in comparison.

It took me 10min to search the term “software” within each group to find a gold mine of Facebook related software tools and people’s opinions of them.

I made a list of all the tools mentioned, in Notion, and visited each website to see what I’m up against.

Honestly, this made me sad. The tools I found were really impressive and clearly run by companies with many employees. It made me feel bad about AdSlicer and what it could realistically achieve.

Curiously, none of the tools I found was solving the problem AdSlicer was focusing on. Jackpot right?

Wrong!

I believe this to be the fundamental issue with AdSlicer. It’s trying to solve the impossible problem of helping you improve the Facebook ads of a potentially bad product.

Bad ads don’t matter as much when the product is selling well.

Good ads don’t matter at all when the product is selling poorly.

Damned if you do and damned if you don't. Case closed.

So what now? Do I just give up and find something else to build?

The reason why I started working on AdSlicer was to help me analyze the Facebook ad results for my (failed) ecommerce store. Then it became about helping my partner optimize one of her client’s Facebook ads for their (struggling) ecommerce store.

I think I’m biting off a little too much to chew, as a novice solo founder, by trying to invent a new product.

I’m not giving up on AdSlicer. Instead, I’m going to compete in an existing underserved market with my own take on how to solve a problem. I hope you like Facebook Ads sales attribution!

Time

I did indeed spend more time not-coding than coding 💪

A report of the time I’ve tracked last month, using Toggl. A report of the time I’ve tracked last month, using Toggl .

This month’s goals

  1. Use Facebook ads to test AdSlicer’s new angle (sales attribution).

    It’s quite risky (time-wise) to update the landing page when I have no inbound traffic. I never know if the copy is bad or if I just don’t have enough eyeballs on the website. This will obviously cost me money, but it’ll be worth it if it steers me in the right direction.

  2. Prototype AdSlicer for sales attribution.

    It would help me sleep at night knowing how I would technically build a solution.

  3. Revive my ecommerce store ( Selv ) in time for winter, by running a 2nd round of Facebook ads.

    This would mostly be handled by my partner. We already took a few lifestyle shots to upgrade the website’s look & feel a bit.

Thanks for making it this far. If you enjoyed this update, it would personally help me out a lot if you could retweet my post:

Alternatively, you can leave your email address below and I'll email you my next monthly update: