Home > Paid Traffic Sources > eCommerce

E-commerce via Facebook (14)

02-06-2017 10:18 AM #1 likebookis (Member)
E-commerce via Facebook

So I have decided to move on to e-commerce. I have found a really nice supplier of watches in China. I have some experience regarding Facebook marketing, however I was wondering if some of the experts here on STM, could give me a piece of advise. I am looking for some sort of a routine to optimize my facebook ads. From scratch. Should i start testing out alot of different targets (age, interests, etc) and then test the banners afterwards? And do you have any tips regarding this.

Really looking forward to hear from you guys. If this thread is somehow placed the in the wrong place, admins, feel free to move it

BTW: For now my focus country will be Denmark since i live here and i reckon my buyers will be males 18-25, however it is always nice to test, test and more tests. Atleast that is what i read on STM forum.

Kind regards

Mikkel

02-06-2017 10:53 AM #2 caurmen (Administrator)

Good question, and well done for thinking about this.

The problem with testing a bunch of demographics first (age, interests, etc) in a very limited way is that you have no idea if you're using a shitty banner to test on them

I'd strongly recommend making a list of demographics that you think might work, then trying reasonably hard to crack each one based on your estimation of how likely they are to want the product. By "trying reasonably hard", I mean going through the usual optimisation process: multiple ads, multiple angles, multiple landers, etc.

If you can't crack a demo after a bunch of optimisation, move on to the next one. But try hard and test multiple approaches - that way you can be sure that it's the demographic that's not working, not just a single ad.

02-06-2017 11:03 AM #3 pain2k (Veteran Member)

Intersect the interests first to have a great list of people really interested in watches to test the product and optimize for website conversions. When you are getting enough sales build LLAs from a custom audience of your view content.

Example: People interested in <niche> magazines are normally a good start. Then intersect people interested in this and that magazine till you have a good audience size. This will be your buyers audience because they've shown affinity to more than one mag.

02-06-2017 12:52 PM #4 stickupkid (Senior Moderator)

Quote Originally Posted by pain2k View Post
Intersect the interests first to have a great list of people really interested in watches to test the product and optimize for website conversions. When you are getting enough sales build LLAs from a custom audience of your view content.

Example: People interested in <niche> magazines are normally a good start. Then intersect people interested in this and that magazine till you have a good audience size. This will be your buyers audience because they've shown affinity to more than one mag.
Not sure/Don't think - people following certain Watch magazines/blogs are interested in; "I have found a really nice supplier of watches in China."
I suggest to target broad male audience with these amazing Chinese watches!

02-06-2017 05:36 PM #5 likebookis (Member)

Hi Caurmen, thank you for the great answer! I get your point, and I will try reasonably hard then!

The standard optimisation progress for facebook would be creating a ton of different groups with interests as well as location as well age and sex ofc. I would also start the targeting the most obvious interest the first, and then try to make my way to find more gold mines out there.

Once again, thanks for the answer :-)

02-06-2017 05:38 PM #6 likebookis (Member)

Hi Pain2k. Great point! So you would basically try and target the most obvious groups the first and then test everything possible. Pretty much the angle, banners and ofc conversion on the websites. Once done, I should move on the new potential groups?

Thanks for the answer!

02-06-2017 05:39 PM #7 likebookis (Member)

thank you! Would you then target a massive group of males and test out angles and banners the first? Then move down to the cheapest targeting solutions, and make profit from there?

Looking forward to hear from you

02-06-2017 06:05 PM #8 pain2k (Veteran Member)

Optimizing your store for conversions is half of the process also. After your pixel is mature with enough sales you go wide open. You don't need a ton of ads to test. 1 adset per interest group at $5 - $50/day on budget. Over that you may have to look at manually bidding. FB algo is a beast.

An interest group would be stuff like Magazines, Competitor brands, Blogs/Forums, Clubs/Associations. Stuff like that. Intersect more than one of these initially for very good quality traffic. So if you were selling pet food and the interest group is brands. Intersect two or more brands initially. Most likely a person who likes two or more of the pet food brands would be really into pet foods.

As a sidenote - This is a proven technique from 6 figure ad-spends used to launch and test products all the time.

02-06-2017 06:08 PM #9 apeek22 (Member)

Nothing to do with targeting but a few tips on picking products.
Use terapeak to research ebay listings ->freaking awesome tool.
Research on Amazon to see items with high(er) number of reviews..

02-06-2017 06:12 PM #10 pain2k (Veteran Member)

Yea Terapeak is great Also recommend it.

02-06-2017 06:46 PM #11 workabck (Member)

Use terapeak to research ebay listings ->freaking awesome tool.
Research on Amazon to see items with high(er) number of reviews..

02-07-2017 06:34 AM #12 likebookis (Member)

pain2k, thank you so much for the detailed answer. Lookin forward to test it out. Have a nice day

03-02-2017 08:42 PM #13 stian (Member)

Quote Originally Posted by pain2k View Post
Optimizing your store for conversions is half of the process also. After your pixel is mature with enough sales you go wide open. You don't need a ton of ads to test. 1 adset per interest group at $5 - $50/day on budget. Over that you may have to look at manually bidding. FB algo is a beast.

An interest group would be stuff like Magazines, Competitor brands, Blogs/Forums, Clubs/Associations. Stuff like that. Intersect more than one of these initially for very good quality traffic. So if you were selling pet food and the interest group is brands. Intersect two or more brands initially. Most likely a person who likes two or more of the pet food brands would be really into pet foods.

As a sidenote - This is a proven technique from 6 figure ad-spends used to launch and test products all the time.
These are some really great tips!

Are you running traffic from FB straight to your product pages or do you use a pre-sell page first which links to product page?

I have found that FB traffic is so expensive (using Optimize for Website Conversions, Pay for Impressions, Auto Bid) that it is really hard to make a positive ROI from the pre-sell approach.

Stian

03-04-2017 01:09 PM #14 pain2k (Veteran Member)

Use a funnel mate.

Home > Paid Traffic Sources > eCommerce