AdWords
1.5K members online now
1.5K members online now
Use AdWords conversion tracking and reporting to measure your results. Have a question about Google Analytics? Post it here, on the Google Analytics Community!
Guide Me
star_border
Reply

Multiple Conversion Types of Different Values

Visitor ✭ ✭ ✭
# 1
Visitor ✭ ✭ ✭

Hi all

 

Just need some outside perspective for a moment.

 

  • I'm on a freemium marketplace - to place an ad you need to sign up.
  • The site caters for 3 main user types - private sellers, private buyers and corporate sellers.
  • Each of these user types can buy adverts across thousands of categories.
  • Main conversion types are: Signup, Ad Placement, Upgrade ($ purchase), Leads (i.e. free download).
  • Money is earned when customers upgrade their profile and there are different upgrade values depending on user type and advert size.
  • Conversion rates from click to signup, and from signup to upgrade vary by advert category.
  • Average value (upgrade price) also vary by advert category.
  • Once a user signs up, they enter an email cycle during which we promote upgrading.
  • At present the site has standard AdWords Conversion Tracking and Google Analytics tracking snippets set up.
  • The AWC snippet tracks signups, leads and upgrades.
  • AdWords is monitored and optimised via AWC (which shows vastly different performance to GA - but is to be expected and is another question in itself).

 

They don't necessarily want to optimise toward just signup value as this potentially undervalues particular advert categories and overvalues others. Likewise, they don't want to optimise toward just upgrade revenue as this would reduce the number of signups in some categories - and the site fundamentally needs signups. At the same time, optimising toward both is seen as double counting.

 

One person believes the best way to optimise is toward upgrade ROI.

Another thinks it should be primarily toward profit, but where upgrade ROI is below target they should regress and optimise toward signup volume multiplied by lead value.

 

My questions are:

 

  1. Whether they need to consider other optimisation options (and what those would be).
  2. What the optimal AWC tracking snippet would look like as a result.
  3. How to handle the transition to managing via Analytics (which shows poorer performance)

 

Thanks, and apologies for length...of this post!

 

 

3 Expert replyverified_user

Re: Multiple Conversion Types of Different Values

Google Employee
# 2
Google Employee

Hi Dan

 

This is quite a complex issue! I think the Community can add some value here, so I'll only add a few thoughts for now. 

 

From what you wrote, it sounds like there's a little conflict or confusion in your team over what your main goals are. It will be more difficult to optimise your AdWords account while you're not too sure what your main goals are. You could perhaps have different campaigns tailored towards the different goals? 

 

In terms of other optimisation options, a couple of ways you can try and optimise your account is by adjusting your bids in light of your conversion data - Enhanced Cost Per Click bidding here I'm thinking could help you gently get towards your optimisation goals. 

 

In terms of the AdWords Conversion snippet, the code you generate in your account is the only one that Google would be able to give you I'm afraid. If anyone else on the Community has any experience with customised conversion tracking code, please feel free to share that!

 

Thinking about the move from using conversion tracking to analytics, as you rightly pointed out there are a lot of differences between these two different ways of measuring account performance. On the whole Analytics gives you more information, as the conversion tracking code simply measures when a page loads and the code fires. As you have quite complex needs by the sound of things Analytics might well be able to serve your needs better! In terms of moving between the two understanding the differences between Analytics and Conversion Tracking might be useful. 

 

If anyone from the Commnuity has any ideas, please dive in!!

 

A.

 

 

 

Re: Multiple Conversion Types of Different Values

Visitor ✭ ✭ ✭
# 3
Visitor ✭ ✭ ✭

Thanks Alys

 

I'm certainly hoping there are people on here who can propose solutions, even from a communication standpoint. I think perhaps people may have been put off by length of my initial post. Don't really know how I can paraphrase tho.

 

They're already trialling multiple AdWords auto-bidding features, enhanced CPC, CPA (Max and, in the past, Target). Top level performance is solid with them hitting upgrade ROI objectives. Agree whole-heartedly that whilst there's this confusion / indecision, maxing the potential is unlikely.

 

Another, more complex, option I'd considered was to calculate signup revenue by individual category and user type. With thousands of categories it'd be a lot of work to run the numbers, and would perhaps need something implementing like a unique internal tracking code per category / user type. Theoretically, from this we'd be able to avoid blended average signup value scenario. It feels doable but brings us back to question of whether this is possible within the AWC snippet.

 

 

I'll keep this topic open for a few days, maybe post a link to it on G+.

 

Dan

Re: Multiple Conversion Types of Different Values

Top Contributor
# 4
Top Contributor

Hi Dan, as Alys has already said, I think part of your problem at least is poor definition of goals, but let's put that to one side.

 

Although it may seem different, I don't think this problem is actually far removed from a typical ecommerce store with a wide range of products.  You have product categories and individual product ranges within those categories.  Some have high values, some low, some convert well, others don't.  In a store situation I'd be analysing the products that can be sold at a profit with AdWords and junking those that can't - so in your case I'd simply stop advertising the Ad Categories that didn't turn a profit.  However, I'm sensing that you can't do this - that you must have regular sign-ups across the board for the whole model to work, is this the case?

 

If so, the only thing you can sensibly do is aim for the traditional AdWords targets.  Better CTR, better Quality Score, better conversions - whatever they are.  I don't see that there necessarily has to be any conflict between optimising for sign-ups and optimising for upgrades.  Although you say the upgrade rates vary, presumably that variation is consistent so if you improved sign-ups across the board the upgrades would increase across that same board by their given rates.  Personally I'd also be interested in why the upgrade rates vary.  

 

So, in short (for me), I say your process could be quite simply to operate on a traditional AdWords optimisation process.  Improving CTR and conversion rates by working on AdWords and the site itself, rinsing and repeating, with a loose overall objective to improve sign-ups, while simulataneously attempting to improve upgrade rates.

 

Is this helpful at all?

 

Jon

AdWords Top Contributor Google+ Profile | Partner Profile | AdWords Audits

Re: Multiple Conversion Types of Different Values

Visitor ✭ ✭ ✭
# 5
Visitor ✭ ✭ ✭

Cheers Jon, great response.

 

You're right, we need new signups are the backbone of the model. Whilst they don't make the site cash directly, without them there'll ultimately be no revenue generated from follow on upgrades.

 

Still toying with the concept of setting up different internal tracking params for each category to derive more granular signup values, based on signup to upgrade conversion and AOV.

 

Agree it's a standard e-commerce challenge, and it's pleasing to see your advice & recommendations broadly fall into what I've already suggested. Always good to get outside perspective - helps lift the blinkered veil from being too close to the problem.

 

Many thanks. I'll leave open for another day in case anyone else wants to contribute, then marked as answered.

 

Dan

Re: Multiple Conversion Types of Different Values

Top Contributor
# 6
Top Contributor

Hi Dan;

WOOO this is quite interesting challenge....

 

You described quite a few (actually many) campaign challenges you are facing. I agree with Alys and Jon:

  • You need to set the goals (This usually more effectively done in a meeting with management). I had recently a chat with a founder of a startup of a freemium service. They target the upgrades . The number of free downloads (by sign-ups) are irrelevant to the revenues . Free downloads bring exposure and the recognition /PR. They are good if you want to sell the technology, not if you want to make money as a business. This is obviously a strategic business decisions that the higher management has to take.
  • After you have set the goals, as Jon suggested, go back to the basics of AdWords to achieve them.

 

Jon wrote a good article, that I like to recommend:

Where to Start (I) - Before you begin with AdWords

 

-Moshe

Moshe, AdWords Top Contributor , Twitter | Linkedin | Community Profile | Ad-Globe
Did you find any helpful responses or answers to your query? If yes, please mark it as the ‘Best Answer’

Re: Multiple Conversion Types of Different Values

Follower ✭ ✭ ☆
# 7
Follower ✭ ✭ ☆

Complex account.

 

I have similiar problem, not as tough as your's with a worldwide dating site.  The challenge between free members and actual subscribers.

 

If you are not doing it already, I like assigning values for various conversion actions.  Understanding the performance between all conversion metrics and overall conversion value is a good start.

 

I am presuming that your sales are being recorded and tracked via ecommerce tracking.  Therefore you can see the actual revenue that is coming in primarily via the sources in Google analytics.  I again will make the assumption that email marketing is driving large portion of revenue.

 

You discussed that there are main segments; Private buyers, private sellers & corporate sellers.  Being able to distinguish which type of person has signed up at the initial stage would be very beneficial.  even understanding between buyer and seller at the very least.  Tracking at this point as different conversion type - ie buyer goes to thankyou page that is different to seller based on preference filled in during signup.

 

From here, I would have a differently tagged email marketing campaign based on these preferences listed above.

 

You should then start the process to cut through, which segments are actually driving the sales.  

 

This should then dictate how you go about improving targeting performance of both segments and follow up email marketing campaigns.