How I Fucked Up creating a product Ad Jet # 2
Hi, my name is Polina and here I am talking about our startup in the field of marketing automation. How we subscribed to all this can be read in the first post.
Today I will talk about CustDev and how we screwed up while testing the target audience. But first things first.
CustDev
CustDev is testing an idea or prototype of a future product on potential consumers.
The last time I did it two years ago in one of the projects, and to put it mildly, I have dubious memories. I remember to find a potential target client and ask him questions. It turned out that this is not entirely true)
How did you look for people on CustDev?
It was the easiest. I wrote on FB that I am looking for people with relevant experience in creating and promoting my business/project. 7 people responded I asked each of them to suggest someone else with the same problem. So I got about 5 correct and 5 incorrect interviews.
In my first interviews, I had the following list of questions:
1. What business are you in? Do you do it online?
2. Do you have marketing experience? If so, which one?
3. Have you ever set up an ad yourself?
4. Have you ever worked with an agency or freelancer to attract clients?
5. “If setup was easier, would you customize it yourself?”
6. Are there any games, quizzes that catch you?
7. Are there any games/applications that you use to kill time, for example on a plane or in a taxi?
Surprisingly, even on these strange questions, I received feedback. True, in the second interview I had a doubt that I needed to update my knowledge.
And so in order, first the conclusions from the answers.
Whom did I interview: fitness trainer, coworking space owner, natural cosmetics store owner, tattoo artist, online furniture store owner.
Key ideas: 3/5 hired a freelancer. The results of their own and others’ work are evaluated on an emotional level, and only one considers the economy.
Pain: it isn’t clear how to work with this, there is not enough time to figure it out, there is not enough time for analytics.
Thank all the gods, at this moment I slowed down and went to remember how to do a normal in-depth interview. The second batch of questions looked like this:
1. What service/product are you selling? What is the most difficult part of selling your product/service on the Internet?
2. When was the last time you encountered this problem?
3. Why was it hard?
4. What did you do to solve this problem?
5. What did you dislike about the solutions you used?
The result has improved dramatically.
Interviewed: the owner of a startup similar to Uber, a product specialist from a bank (this was a pass, but interesting), a product specialist trying to launch a startup in the field of blogging, a dance shoe seller, a pastry chef.
Key ideas: all the respondents have problems with setting up ads, moreover, regardless of whether they set up ads with their own hands or delegate.
Pain: it is not clear what message to broadcast to the audience, in some sites a/b tests and payment for installation are not implemented (relevant for applications), there is no clear analytics.
So we decided to focus on solving three main problems: determining which sales channel to go to the client, hiding the internal advertising settings (to give the client only the result), and understandable analytics.
Coming to life
First, we made up the landing page on a Tilda. I would like to say, “Oh, it was simple.” But we needed to insert a test there. We did not find anything better than the onlinetestpad platform (it feels like a site from the early 2000s), but we managed to use it.
What was the point?
The client (artisan, small business owner) answers our questions and at the end receives a marketing strategy, in total we had about 36 different answers. For example, if you say that you are selling a product and you have a website, but you cannot buy on it, then first of all we advise you to go to marketplaces, and secondly, add the function of selling goods to the site.
We decided that the landing page as a whole was ready and launched the target. (Not so easy)
Test № 1
Platform: FB, Insta
Region: Estonia, England
Target audience: freelancers, m / f, 18–54
In both cases, men responded to our ads better than women. In Estonia, the cost per click was cheaper. We collected a database of 50 freelancers, talked to them, and found out that they are ready to work with us if we have an automated part (like on prototypes).
Of course, part of it twisted at the temple and did not understand why we were doing advertising at all, but in general, we received feedback from the market. They did not take pens to do it. Each freelancer has an average of 5–6 clients and we physically and financially cannot meet such a volume of work.
Test № 2
Platform: FB, Insta
Region: Estonia, England
CA: small business, m / f, 18–54
And here we were disappointed. We did not change the site and launched advertising directly to it for small business owners. Of course, small businesses did not understand at all what we want to offer and why they need it, we just drained the advertising budget.
At this stage, we realized that we had made a mistake with positioning and we needed to change something. We wrote the text and inserted what we thought were “understandable prototypes”, but we completely forgot that not all representatives of our target audience understand what we mean. And we need to explain the product more clearly. And what we came up with I will tell you next time. This is how our startup is doing.
