Home > Paid Traffic Sources > eCommerce

---

Switching fb conversion events from View Content to "Add to Cart" (11)

---

I have read the facebook web conversion guide which states that a minimum of 15-20 conversions are required to optimize for this event. If you are not getting the conversions then:
"Start out by optimising for a frequent conversion, even if it's not the one you ultimately care about, to get a sense of how conversion-optimised delivery works. Then, once you feel like you understand, try optimising for the one you do ultimately care about, even if it's rarer."


I wasn't getting conversions with add to cart. Therefore, I have switched to a "View Content" web conversion ad and I have gotten 15 conversions. I want to now switch to add to cart.


A facebook rep where I asked the question on fb help community told me to:


>>> Duplicate the successful View content ad and change the conversion to Add to Cart, as too many edits to an ad can cause your ad to not optimize properly.


But here's my question which they were not able to answer:


If you duplicate the adset then doesn't the optimisation data get lost? I thought the optimisation took place at the adset level and so by duplicating the adset any optimization is lost?


If an ad is started with AddToCart conversion event and nothing is touched on the ad and it runs for four days (while collecting enough conversion events), is the best approach to then edit this ad and change it to Purchase conversion event?
This would then surely keep the optimisation data? It really isn't clear on the help pages and forums on the best approach here and there's conflicting information!


any help would be appreciated.

---

For what it's worth, we started directly with Purchases as a conversion goal - as you said, optimization happens at an ad set level so even if you duplicate and keep the same conversion goal, all that is lost.

I think the idea you explain is more about getting some confirmation that the audience has some sort of interest in your offer and also perhaps you see how ATCs and Purchases look when it's optimized for VC.

---

Yes, you're right to say that! If you started off directly with Purchases as conversion goal, how do you get confirmation that the audience does have the interest in your offer, and that your offer is proven to convert?
I use ATC as a conversion goal first because I need to get 15+ conversions for FB pixel to optimize, but if I start off with Purchase, and if I don't get any conversion, then there's nothing for the pixel to optimize

Or do you just start with Purchase as conversion goal and if there's no conversion, you just move to a new product?

What are your thoughts on these?

---

The people I've talked to recommended going for purchases and paying attention to the cost per visitor and conversion rate. If you don't have conversions, it's not like Facebook stops sending traffic. Somehow, for me it was never the case that there were 0 sales. Biggest issue at the beginning was high CPM and low CTR - we used boring ads with bad targeting. Basically the combination of the 2 stats told me that I will not be able to profitably sell items in my price range.

It can be more than your product why people don't buy, for example a less than optimal product page can severely damage your conversion rate, aka sales.

I think you have to do a bit of back and forth for optimizing both the audience/targeting and the product page. The good thing is that you can get a lot of the product page done properly before running traffic since some main tips apply everywhere.

Also, I'm learning, I'm not an ecommerce expert by any means - that said, I'm not scared money so I test a lot of my assumptions with paid traffic directly.

Don't forget the amount you spend vs the product price though, it's like AM here that you can't just buy $5 worth of traffic and expect to sell a $50 item without any optimizations. I don't want to give rules of thumb when I'm so new, but I would warn against drawing conclusions with only 1x sale price spent on an ad set if you don't have the product page well optimized already for example.

---

How did you realize that it was because of your boring ads or bad targeting?
I'm assuming you realized that because you did plenty of split tests to test the ad/images/targeting/etc

However, if you did create many different split tests, do you carry out the split tests at the ads level or the adsets level?
I'm asking this because as you said, the pixel is optimized at adsets level, so I'm assuming you do all your split tests at the ads level?

And if that were the case, then how do you split test targeting? (Since to split test it, you'll have to create another adset)
And when you create a duplicate campaign to split test targeting, you start off with a "fresh" pixel again for that adset, and won't that make the results skewed?

Hope to hear your insights on this

---

Ad set level always

Realized from feedback (from people already running profitable stores), results (of different ad sets), and research (into competition).

Products we were promoting are all top sellers, and usually promoted with different ad types than we tried.

We didn't make a classic split test, the kind you can make on pops for example - but we launched several ad sets at once with different targeting and ads combination (one ad set for each) and we also looked at CPV (cost to get a visitor to the product) because there is a range you can work with based on your product's price. If you sell something for $2 and it costs you $0.75 to get a visitor, then $1 for the product... You don't really have a chance at making money, right?

It's also much less black or white than CPA - I've said this before when talking to others about CPA marketing: The focus should be on improving your previous results all the time and not simply get profitable. Step by step improvements are what get you that huge scale, it's not gonna turn from -90% ROI to +25% ROI instantly.

---

hi

if you have an adset that is set to view content but you see that you are getting enough ad to carts for the pixel to optimize, you can just switch events in at the adset level and it should start optimizing from there for ad to carts...

the point of starting with viewcontent is that you can rapidly acquire the necessary data to create a lookalike audience since people are more likely to view something then they are ad it to a cart.

then you duplicate the adset and use the ad to cart event with your view content lookalike audience and you are essentially transfering the data you have aquired.

also you might want to keep your view content adset running if its break even or at a slight loss as it will be gathering more data and getting you a better view content lookalike audience that updates over time.

then you do the same thing once your ad to cart adset gets enough data to create an add to cart lookalike.

once again you duplicate your adset and set it for website purchase event and use you add to cart lookalike audience.

eventually, you create a lookalike audience from your purchases and create an adset with that audience and objective.

essentialy u are hopping from one conversion event to the next and using the easiest data to gather first to get you to the end goal of a very optimized purchase event lookalike audience.

You could just start with a purchase conversion event, but since these are the least likely events to happen, it will take a long time to get the data you need and will cost more money to figure out whether that adset can be profitable, especially if your targeting isnt perfect.

Once you have a view content lookalike audience you dont have to worry about the correct targeting anymore since facebook already knows who is most interested in your product and will start optimizing for you and getting more precise over time, and the quicker you get to that point the better.

Atleast that is how i understand it. someone correct me if im wrong!

---

The issue is that people who click ads are not always buyers. I've ran into the issue of having many views and no sales, and that doesn't make a lookalike audience valuable.

Like you said, you still rely on the audience to buy, even if your objective is set to something else.

It makes sense that people who view your content could be more likely to buy though, so then you are basically trimming down from a bigger audience to a smaller one until you target buyers only. I would not think it should be a profitable ad set from the beginning with VC goal - pretty sure an ad + product so hot will be profitable with sales goal as well.

---

I should have specified that obviously, you would only do this with view content adsets that are showing some initial promise

if its a dud adset then just chop it!

but if its the kind of adset where your cost per purchase is close to break even, then this is the time to let it run until you can create a view content lookalike audience.

Over time as the pixel gathers data and you move up the convertion event ladder, you should start turning a profit as you start to really target those that will purchase the product.

for me this is the way you optimise a promising adset that is losing you money into one that can make you money.

but again i would only do this with adsets that are already showing multiple conversions and close to break even.


.

---

Doesn't FB optimize step by step anyway though, if you have no previous data?! It first wants to get some clicks to your page, then the other conversions?

It sounds OK on paper, but I think either FB updated already, or will update to make its algorithm smart enough to do this process by itself instead of having to rely on you, the advertiser to correctly adjust conversion goals in future ad sets.

Anyway, I've heard this method presented a few times and what I can see is that it works also with using directly Website Purchase as a goal and keeping an eye out for all the numbers to be in line with product price.

I haven't tried the step by step method but I don't see a reason why it wouldn't work - it should be better than direct to purchase, at least from a mindset perspective, that you can measure your progress in smaller steps.

---

The main advantage is speed and money spent to get to the point of creating lookalike audiences. you just spend a lot less time showing ur ads to people that might be completly disinterested and waste money.

it will take a while to get enough people to purchase to create a lookalike audience

and if you are working with just interest targeting, it will still be very broad and will take a while to get that data

to create a lookalike audience from view content can be done very fast and will allow you to start focusing in on your true target audience quicker.

if you just go straight for the purchase event with interest targeting you might have your ads thrown up to people that are completly disinterested for a long while before you get the data you want to then create that purchase lookalike.

maybe facebooks pixel is smarter now, im not sure. but the fact that they havent switched over completly to just a purchase event and not done away with view content and ad to cart events is probably a sign that it isnt that smart just yet.

anyway this is just the way i do it at the moment. but i do sometimes just go for the purchase event from the get go if i feel the item im pushing as very obvious targeting.

---

Home > Paid Traffic Sources > eCommerce