Question

So based on advice I got here and a little bit of research, I decided to switch to a SPAG(Single Product ad group)(Yep, it appears that's a thing) approach for my shopping ads. I used ads editor create an ad group for each of my products.  It looks like I will have to add products one by one to each ad group. That's too much work considering I have 2500 products. Is there a quicker way to do this?  

Answer 1

Yes, I used JavaScript and made it spit out a csv thru which I import it to g ads editor

Comments

Tell me you made a youtube video doing this?  

---

@overrate9576 unfortunately I do not, I’ll be surprise if a video does not exist for this

Answer 2

Let's assume that you will do this. Just a stupid question: How you will manage them?  

Comments

I know a question coming from you is not going to be a stupid one. First time splitting it this way, so I will be taking one step at a time. If it doesn't work, reverting to my usual method is quite simple. Having said that, I would run it more or less the same. Create 2 or 3 campaigns, different priorities, use negative keywords to sculpt traffic etc. Just the fact that I will get to see the exact product triggered by search term, seems like really valuable insight to have.  

---

I read your article earlier today. Ranks quite high on google search. I would just prefer to get this insight within the ads platform. This attempt might fail, but I've already learnt so much just today, makes it worth trying. After that, I'm going to try your way as well. Because why not.  

---

@overrate9576 I hope you will have success, even if I think that it's not possible for a human to analyze such volume of data and the most important: to get decisions. I suffered just like you in the past.  

---

@raynell9 the bigger problem is that you can only put one product on a SERP. The real fun in Shopping is when you take over the whole box.  

Answer 3

Parabola. io if you don't know JS

Answer 4

With that many products, i'd optimise at bid level for each product based on say last 30 - 60 days data and return on ad spend. One campaign You can always pull out the top 20% and move them into their own campaign so you can adjust the chedule, device geo etc. per campaign. I've had consistant 8x roas over a number of years doing that. Five figure ad spend pm.  

Comments

Pretty much what I usually do as well. However, with this set of products, the irrelevant traffic I am getting is insane. I'm adding negative etc, but trying to see if I can do some next level shit here

---

How do you determine what negatives to add?  

---

I have a list from my other campaigns in the same niche. That's what I started with along a few pulled using keyword planner. From them on, I'm just adding negatives using the search term report. Anything else you think I should be doing?  

---

Search term report only - at least 100 clicks. Reduce the bids for anything under target roas appropriately. Eg. Target roas is 6x if the current avg cpc over the last 30 days or so (whatever makes sense data wise) is 50c and roas is 3x set the bid to 25c. if roas is 12x nudge up daily to $1. Do it for every item. If you're bid is the same for everything, that's never going to work and you'll be negging the wrong keywords. Tune all variables to target roas. Implement a 24/7 bidding script. Find all items with low impressions and increase bids till you're getting a decent amount of impressions and clicks, then adjust the bids based on roas. Implement a retargeting campaign. Search campaign for your brand. Search campaigns for the brands you sell, if you can ie, they are not Tm. If you have enough conversions, use data driven attribution. Check auction insights - find your top competitors and make sure your pricing is good. Your pricing affects ctr which affects whether your ad is shown or not.  

Answer 5

Don’t do that

Answer 6

I've got a tool for that.  

Comments

What's it called?