Day 26: Adventure in Ads, Lead Growth and Pre-Launch Hustle
In our last update on Day 24, we did a three-part article on how to do sales in startups. Today, we have a lot of housekeeping to do and some updates to provide on the activities we’ve started over the last couple of weeks. It’s been a crazy week and Slobodan is heading out for a 2 week vacation on Friday, so I didn’t have that much time to work on the side-hustle.
We started a Facebook Ads campaign with the help of CofounderGPT on Day 22 and we set it to run for two weeks. I was monitoring the progress of the campaign and we had a clear winner with our A/B test within a few days in terms of cost per 1,000 views. But we did not get any conversions from the Facebook campaign after about almost $80 spent on ads so I decided to pause it 2 days before it ended.
Facebook ran the ads on the weekend as well, even though I’m pretty sure I configured the ads to run only on weekdays. And of course just like Google, Facebook manages to go slightly over your daily budget every day. It’s kind of annoying.
Anyway, we had some success with Vacation Tracker when we ran a Facebook campaign many years ago to find some beta users. This time it didn’t work as well as in 2017. It could be because we did something different in our campaign or it could be that Facebook is not as good of a channel as it was before for this type of product.
I actually always felt Facebook was a bad channel for B2B products and we were never able to get it to work for us at Vacation Tracker after that initial beta campaign. And even that first beta campaign, we’re not sure how many of the people ultimately converted by trying out the product. So for now, we’ll stop using Facebook for direct ads and may try retargeting ads at some point down the line.
Contrary to the Facebook ads, I am happy to report that the Google ads campaign has been a resounding success! we’ve spent $110 over a 2 week period (excluding weekends) and managed to bring 223 people to our website using the keywords and ad copy that CofounderGPT helped us create.
Out of those 223 people who visited our website, 19 left their email addresses. Which works out to about $5.8 per lead. Not bad. The Google campaign performed much better than we expected.
Here are the details:
I’m no ads expert, but I think our click through rate was pretty good for most of these ads. Kudos to CofounderGPT for writing excellent ad copy and for coming up with a good landing page.
There is only one problem with the Google Ads campaign. Most of the email addresses are Gmail addresses. The last time we did a beta campaign like this through Google, we had more company email addresses. Out of the 19 we captured, only 5 are company domains and the rest are Gmail.
So we’ll see how reliable these leads turn out to be but I’m still pleasantly surprised about the success of this ads campaign. we’ll certainly continue with it after we launch the Knowlo beta.
Mailing list and Newsletter
We didn’t provide any updates about this since, so Let’s have a look at how this is performing:
Our newsletter open rate has been declining over the last 3 weeks. We were kind of expecting this with the increased number of emails that have been added to the list. Some of them are probably crappy quality leads or even non-existent emails (we had 4 emails bounce on the last newsletter).
It doesn’t seem to be deliverability issues like last time, almost all the emails were delivered to people who subscribed. So we have to see whether we want to continue sending these newsletters or whether we should just stop when we with a final one when we launch in a few weeks to let everyone that the product has been launched.
This is not a decision I’m going to make without Slobodan, but I will try to work with CofounderGPT to improve the headlines in the meantime. Maybe our headlines have gotten a bit boring and don’t explain What’s in the newsletter. Since we’re going to be launching within a week or so after Slobodan gets back from vacation, there are really only 3-4 newsletters left until we launch and we can continue doing them.
In terms of how long this takes and how It’s put together, I’ve gotten pretty quick with it with the help of CofounderGPT. It usually takes about 1 hour to put together. The process is quite simple.
First, I put the articles we wrote and the excerpts for them into the newsletter. Sometimes I ask CofounderGPT to rewrite one or make it a little more descriptive of the article. Then I ask CofounderGPT to scan through a few “latest news in AI” Twitter threads that I find and save each week and ask for recommendations that are relevant to startup founders, entrepreneurs, product managers and CTOs. I usually already have a pretty good idea of which news we’re going to cover in the newsletter but CofounderGPT helps with summarizing and sometimes picking the articles that go in the newsletter for the AI News section. Finally, I put what we have through CofounderGPT and ask it to write an introductory paragraph for the newsletter.
I always make small changes to the text that CofounderGPT writes to make everything more consistent. But It’s getting better at writing these newsletters because it has all the previous ones loaded as context whenever we write a new one.
We’ve reached Day 26 and we still haven’t invested much time into our branding. And I still think it would be premature for us to do so. I’ve actually gotten a few compliments on our simple Knowlo logo which took me 5 minutes to design on Day 2 using Canva and we’ve stuck with it since.
For entrepreneurs who are starting out, branding can be a rabbit hole which sucks up time and distracts you from What’s important in the early days of starting a company. I’m not trying to say branding is not important. Let me be clear by saying that branding is an exceptionally important part of any company. Just not when you’re trying to quickly launch an MVP and validate whether your product idea even has market.
One thing I’ve observed over the years is that people will use a horrific looking website or app if it solves a problem for them. And since you’re looking for early adopters when you’re launching a startup, they’ll especially be more likely to accept an ugly product if it does the job for them.
The inverse is not true though. Even with the nicest branding and design, unless you are solving a problem for your customers, they will not buy your product. So my opinion is that serious investment in branding (both time and money) should only be done after you have some kind of product-market-fit.
Google Search Console Issues
You may recall that on Day 18 we connected Google Search Console to the Knowlo website. I decided to check in just to see if our website pages are being indexed by Google and to my surprise, only some of the pages have so far been indexed.
As you can see, less than a third of our website pages are currently in Google. Which means the rest of the content is not searchable through the Google search engine. We have submitted our sitemap when we were setting up Google Search Console so it has all the pages of our website.
Here are the reasons that Google gives for why It’s not indexing those pages:
I haven’t been dealing with these kinds of issues in a long time, and Google Search Console has evolved a lot over the years. I needed some help to figure out what all this meant, so I asked CofounderGPT for an explanation:
This was a good starting point but I had to dig a little deeper. What I found was that there are some duplicate pages with www and without www in front of them that are not being indexed. Those are mainly in the pages with that have a redirect problem and from what I see, I’ll need the help of a technical person to help me resolve that issue. There are also some WordPress template pages left over that need to be deleted and some pages that Google just decided not to index.
Time spent today: 4h
Total time spent: 178h
Investment today: $54.44 USD ($34.44 for Facebook Ads and $20 for ChatGPT)
Total investment: $1,284.45 USD
Beta list subscribers: 86
Paying customers: 0
It’s going to be a bit slow for the next 2 weeks while Slobodan is on vacation. Although he left 1 technical update ready which I’m going to publish tomorrow, our MVP development is going to grind to a halt for a little while. So I’ll be working on building our list to try to get 150 people on it before we launch.