Would this be a good strategy for split testing a niche?

Hi guys,

I'm planning on launching a niche e-commerce and I've decided to use my small budget to invest in some FB advertising. First, I want to split test my ads to find the most engaged users, age, location, interests etc.

It would be fantastic to get feedback from this community as there is so much conflicting information online and it can be overwhelming at first.

Overall Test Plan - My budget is $5 per adset, with a total of 8 different adsets - $40. I have two campaign goals, conversion + PPE. In each adset, the only variable that changes is the interest. After I've performed the first split test I was going to analyse the results and then move forward with the ads that performed the best and create new ones based on the top performing demographics, interests and variations of image and copy.

Campaign Objective - Overall I want to SELL and I wasn't sure which campaign objective to go for. So I decided to go for PPE and conversions. This way if one works and the other fails at least I may have potentially built up a small following on my page to later re-target directly, or proceed with the ads that created conversions. Obviously, if neither happened it is pretty screwed straight off the bat and would have to rethink my process completely!

Interests - I've researched the specific interests and found ones with high affinity (over 100), they have at least a reach of 50,000 (ranging up to 150,000) and all fall into the category of "hardcore fans"... people that are more interested in the niche than the average consumers and aren't just there for the funny videos, images or whatever. My plan is to have a different interest for each ad set, whilst keeping all other variables the same. I also only want to use 1 interest per ad-set so I know which one will be the cause of the high performance, as opposed to stacking lots up and not being able to determine which one makes the most difference.

Images and Copy - I'm keeping the copy and images the same for each set, however, I'm using two different forms of copy across six different images as to test which variation of the image and copy works the best. My understanding is that by keeping these the same it will help me determine my best-performing images and copy combination. (The attached excel doc may explain this a little better.)

Gender - I've researched my niche (Audience insights) and have found that in each interest around 80% is female based, so I was going to keep it simple and go straight women in all my ads. I know nothing should ever be assumed when running tests but surely with these figures being so high it validates that thinking, no?

Age - For each interest, I looked at the demographics on the audience insights tab and picked the sectors with the most activity, normally 15-20% and around 2-3 sectors per adset.

I think that pretty much covers my strategy, I have literally no idea if it's too broad, too narrow, overly complicated etc.

So, if anyone out there fancies reviewing this and giving me some guidance and feedback that would be absolutely amazing. Be as ruthless as you want in your feedback, after all, I'm here to learn!

If you need more info or have any questions I'll do my best to answer ASAP.

THANKS!

(Excel doc is attached.)

 

I realise the image may be too small to fully understand the testing. It goes like this - (Top to Bottom)

• Campaign Objective (PPE + Conversion)
• Adset
• Copy
• Images.