{{ searchResult.published_at | date:'d MMMM yyyy' }}

Loading ...
Loading ...

Enter a search term such as “mobile analytics” or browse our content using the filters above.

No_results

That’s not only a poor Scrabble score but we also couldn’t find any results matching “”.
Check your spelling or try broadening your search.

Logo_distressed

Sorry about this, there is a problem with our search at the moment.
Please try again later.

I hate the word granular. I spend half my life tweaking and fiddling with PPC campaigns across different platforms.

The word granular invariably means spending even more time setting up campaigns. The problem is, the only way to achieve, monitor and maintain success in PPC is by going granular. The same holds true for Google Product Listing Ads.

I first went granular with my bog standard AdWords search network text ads soon after starting out in PPC. I went granular with my Product Listing Ads at a much later stage however.

When I first setup Product Listing Ads (PLAs) I had to do so with the assistance of the official Google documentation and a few third party guides. It’s all a bit fiddly.

Most of these guides seemed to encourage large ad groups for one reason or another. Against my better judgement I just went along with it, wasting thousands of pounds in the process.

Having worked on various PLA campaigns I can tell you in no uncertain terms that you must go granular if you want greater control and better results.

If you don’t believe me, just try lumping together massive ad groups for a few weeks and see how it goes for you. There’s simply no alternative to going granular if you want to see a positive ROI whilst generating data that can be worked with in order to improve the performance of your campaigns.

Shock, horror

What prompted me to go granular was the fact that my PLA spend was almost identical to my regular text ads spend, but my conversion rates on PLAs were less than half that of my regular text ads.

Whilst going granular would have little effect on my conversion rates, it would allow me to work out which products just weren’t converting, and which ones were.

Consequently I’d be able to improve the landing pages for low converting products – or axe the ads altogether and divert the marketing funds to other, better converting items.

The way I’d setup the campaigns was to lump all items in a certain category in the same ad group. PLAs don’t use keywords to trigger ads. Instead you define the parameter on which Google groups your PLAs and Google decides when to trigger that ad. You supply Google with a product ID, product type, brand, condition, an AdWords label, or an AdWords group for each ad group that you create.

Google then decides when to show your ads (usually when someone searches for a keyword contained on the product landing page).

The furthest you can drill down in the list of variables is product ID. By creating an ad group for each product ID you’ll be able to obtain metrics for the performance of each and every product on your website.

The other options cater for larger, broader groups. Analysing the data and improving your ad groups can be hard when working with broad groups.

Obtaining and using data

I spent a few days beavering away on my revised ad groups. Each ad group targeted just one product ID. That meant for the 300 or so products in my ecommerce store, every single product was in an ad group of its own. That’s 300 different ad groups across a couple of campaigns, now that’s granular!

A few days after configuring my ads in such a way, the results already started to show. One of my previous ad groups contained ten different types of a specific item. Nine items in the range were under £7 in price. One particular item cost £22 however.

Prior to using product IDs to configure my ad groups, all of those items were in the one ad group running with the same bid. The group was just about breaking even in terms of ROI.

After analysing the data from each individual product ad group I was able to determine I was bidding too high for the cheaper items, and too low for the higher ticket item.

I increased the bid for the higher ticket item ensuring my PLA was always shown on the first SERPS page. I made more sales and because the item was more expensive, I could afford to bid three or four times more than I could for keywords pertaining to the cheaper products.

This pattern emerged across most of my inventory and categories. I can now load up my AdWords account and see a product by product performance report. Those PLAs that perform well have their bids increased, those that perform badly are tweaked whilst I try to improve their metrics.

Sometimes I come up against a brick wall – a product where no matter how hard I try, I just cannot make it profitable.

In such cases I simply pause the ad group and move on. Don’t be afraid to do this in your campaign. Often I’ll revisit the ad group a few months later and have another crack at making it profitable – with varying levels of success.

Fixing things

I look at conversion rates very primitively. A high conversion rate means your ad group is good. A low conversion rate means it’s broken. There’s no point me throwing around different conversion rates as a benchmark here, every website and every niche is different.

In some niches I might be over the moon with a 10% conversion rate. In others a 10% conversion rate might not even get me out of bed in the morning.

Here are five things you can do with the data generated by granular PLA groups:

  • Isolate best selling lines and increase your bid on those particular products. 
  • Decrease the bids on lines that generate lots of clicks but few conversions, allowing you to spend more money on traffic to products that actually convert. 
  • Work out your PLA marketing ROI for each and every product in your inventory. Products with a poor ROI or with low sales figures can be discontinued to allow you to focus on more popular items.
  • You can spot trends that might help with ordering inventory. Sometimes a product might be mentioned on the television or in a newspaper – search traffic goes through the roof and you have to make a special order with suppliers to keep up with demand. 
  • Use impressions data to determine which products you push hard in terms of search network text ads and organically. If you can have three links to your website on the first page of the SERPS for each and every product you stock, the chances of users clicking through to your website increase exponentially compared to if you have just a PLA on there. 

Final thoughts…

So many people out there are lazy and they throw loads of keywords or products into the same ad group then wonder why they see poor results, or why they can’t seem to establish patterns in the data they gather.

Using product IDs to laser target ad groups on the PLA platform won’t allow you to automatically see great results, but it gives you better data to work with, data that’s relevant to every product you stock, not just generalised data that covers large swathes of your inventory. 

I’ve seen people making the argument that you shouldn’t set PPC bids based on profit margins on individual items. Whilst I agree with this to an extent it’s something you need to watch very carefully.

If you’re working to a £1 margin but CPCs are 50p and your conversion rate is 10%, there’s really not much point – you’re making a big loss on every sale.

Without drilling down and putting individual products in their own ad groups it’s impossible for you to determine whether bidding for certain products is worth it.

Overall you might be scraping a profit, but if you remove those poor performing items from your campaigns altogether your profits can improve dramatically. You’ll spend less money on PLAs burning up a large amount of your budget for no profit – you’ll then be able to take that money and spend it on more profitable ad groups, bringing in even more sales.

Take my advice and go granular with PLAs. The main disadvantage is the time it takes to setup the ad groups. The advantages can’t be articulated though, there are too many to list.

Avatar-blank-50x50

Published 4 December, 2013 by Nick Whitmore

Nick Whitmore is Managing Director at Whitmore Media Ltd and a contributor to Econsultancy. You can follow him on Twitter and Google+

3 more posts from this author

Comments (2)

Jonathan Beeston

Jonathan Beeston, Director, New Product Innovation, EMEA at Media & Advertising Solutions, Adobe

Great post Nick. No one would advocate putting all their keywords on broad match, the same is true for PLAs. Granular structure brings control.

You have to run on PLAs and conventional text ads together. Understanding the balance between the two is critical.

over 2 years ago

Avatar-blank-50x50

David Rothwell

Granular is definitely the way to go.

You also get great insight into search queries, and how well G is matching your Product Targets with user searches (from what I see, they are incredibly good at this, but it will depend on the quality of your data feed).

I've gone as far as using an excel spreadsheet to create single ad groups and product targets when a client had 140,000 product ID's (yes, that's one hundred and forty thousand - he was an affiliate). I had to spread it across multiple campaigns as we up against the max ad groups per campaign limit.

Why? Because I could, and I wanted to see what would happen. It didn't actually take that long either. We immediately got more impressions as a result.

You can customise a promo ad to every product this way too. There's lots of advantages to it.

Thanks for the post!

over 2 years ago

Comment
No-profile-pic
Save or Cancel
Daily_pulse_signup_wide

Enjoying this article?

Get more just like this, delivered to your inbox.

Keep up to date with the latest analysis, inspiration and learning from the Econsultancy blog with our free Daily Pulse newsletter. Each weekday, you ll receive a hand-picked digest of the latest and greatest articles, as well as snippets of new market data, best practice guides and trends research.