Google Ads: Building audience strategies to improve your targeting and performance

Initially absent from Google Ads (formerly Adwords), the concept of audience targeting arrived in the advertising network in 2010. The arrival of Facebook on the advertising market in 2007 with the promise of being able to target individuals via their profiles prompted Google to follow suit. Until now, Adwords offered advertisers the option of targeting search terms on its search engine or locations by affinity on its AdSense inventories and its Display network. In 2010, the first audience offered on Adwords was that of the advertiser's own site, enabling the advertiser to remarket to Google.

Almost 10 years later, Google Ads has developed new possibilities. How can these audiences be exploited? How can campaign performance be analysed using the notion of audience? What strategy should you adopt?

Let's find out what's at stake in search engine marketing (SEM) as part of a Google Ads audience strategy.

See all software Search Engine Marketing (SEM)

What is an audience?

An audience is a profile of users based on one or more criteria.

Every second, a large number of users carry out searches on the Internet or browse sites/applications. Building an audience means providing an answer to the question: " Who are they?

Google uses multiple sources of data to profile its users, the main ones being :

• Search history,
• Gmail email,
• Browsing history (via Chrome or the Google Display Network),
• YouTube history,
• Google Account activity.

The use of audiences in Google Ads allows you to qualify the profiles of users who are likely to be exposed to your ads.

Grouping your audiences

The best practice is to classify users into 3 main clusters:

• Customers: Already users of your services, they present a potential and retention challenge.
• Hot" prospects: This is your "incubator" where we classify users who have interacted with your services but who are not yet customers. This can also be referred to as a lead nurturing strategy.
• Prospects: Users who are a target for your services and who have never interacted with your brand.

Today, at JVWEB, an e-marketing agency specialising in search engine optimisation, we estimate that advertisers miss out on 20-30% of their conversions if they do not use an audience strategy.

Looking at the subject from another angle, the AdWords account performance analysis tool Seiso shows that advertisers can waste up to 45% of their budget on irrelevant audiences (which should be excluded from their campaigns).

The different types of audience

Audiences are determined on the basis of different data: data from Google or data from the advertiser.

Audiences derived from Google data

• Demographic/geographic: Allows you to reach or exclude people who are most likely to be in the same age, gender, parental status or income bracket.

• Affinity: Allows you to reach or exclude people who are actively looking for products and services like yours.

• In-Market: Enables you to reach or exclude people interested in products or services like yours.

• Similar: Find new users identical to those on your remarketing lists (-29% average cost per acquisition observed - source: JVWEB) or similar to centres of interest, or (and this is where it gets really powerful) a Customer Match list.

Audiences derived from advertiser data

• Visitors: Allows you to reach people who have already visited your website. This is known as remarketing or retargeting.
  
  The average acquisition cost for this audience in RLSA (Remarketing List For Search Ads) has fallen by 44% (source: JVWEB).
  
• CRM: Reach people from your CRM database using the qualification elements it can provide. This is known as Customer Match.
  💡 NB: we strongly advise you to provide Google Ads with an encrypted data file (hashed in SHA 256 format), particularly if you use a third party (consultant, agency, etc.) to manage your campaigns.
  
  The average acquisition cost for this Customer Match audience has fallen by 57% (source: JVWEB).

How do you use audiences?

The types of audience available for each type of campaign

Personalised audiences: your audience, created by you

You know better than anyone which users are likely to be inspired by your brand. This is where your business expertise should enable you to control the tool's automation.

The method:

1. Identify your core target,
2. Reach them with a minimum of waste,
3. Measure your performance with these audiences.

You need to think about your audience approach in terms of the level of engagement of your prospects.

Here's a diagram describing how to use the different audiences to reach a broad target while staying on the mass most likely to engage with your brand:

Maximise the conversion of your traffic

Use retargeting in Search and Shopping

To this, you need to add the behavioural dimension: you need to take into account the depth of the conversion tunnel and the level of engagement of the user you are going to target.

💡 NB: The conversion rate for retargeting via the Shopping + Search network is 20% higher than for standard prospecting (source: JVWEB).

In this way, you can create 3 different lists with distinct bidding strategies in order to target users according to their level of engagement in the conversion path.

We advise you to start your campaigns by setting fairly high bids to improve your learning curve and reduce the time between you and the optimisation phase.

Here's a good model for launching your tests:

Using similar audiences

Similar audiences are added automatically by Google Ads based on your own audiences.

Make the most of your CRM data

Segment your database to export as many contact lists as possible.

Then integrate them as Customer Match audience lists in the Google Ads interface.

Audience optimisation strategy

Prerequisites and best practice before optimising

Before you start, make sure that everything you need is in place:

1. A Google Ads remarketing tag is present on all the pages of your site.
2. You have authorised the collection of Data for remarketing in Google Analytics and your Google Analytics account is linked to your Google Ads account.
3. Your audience lists have been created with a satisfactory level of granularity, while maintaining a minimum size for each list (minimum of 100 contacts per list).
4. All your audiences are added to your campaigns in " Observation" mode.

We also recommend that you :

1. Use Audience Insights to identify the profile types of your audience.

Use the audience statistics tool in your Google Ads account:

2. Create duplicate audiences with Google Ads and Google Analytics to maximise the size of your lists.

3. allow time and observe performance when volumes are significant before making decisions (2 weeks to 3 months depending on the level of account activity).

Optimisation method

A Google Ads Seiso report can be used to immediately identify the best-performing demographic audiences and the most problematic ones, at the level of the account as a whole, but also for the largest campaigns.

Here are a few examples:

On this basis, you can plan optimisations:

1. Adjusting your bids upwards or downwards depending on the performance of your audiences (based on ROI or cost per acquisition).
2. Exclude audiences that are not performing well.
3. Offer personalised messages.
4. Use Smart Bidding and feed your data into the Google algorithm.

Finally, think about personalising your messages within the same ad using the "if" function. In this way, you can refine your communication or push tailored offers (for example, first-time buyer discount vouchers for prospects, free delivery for customers, etc.).

From a business point of view, it may be important to incorporate the notion of lifetime value into your audience strategy.

Depending on your business model, a customer may only be profitable after several conversions.

You will find a report on customer value in Google Analytics. Customer value data is available in all Analytics accounts. No changes to the tracking code are required:

1. Connect to Google Analytics,
2. Access your view,
3. Open the Reports section,
4. Select Audience then Customer Value.

Google uses a classic RFM (Recency Frequency Amount) to calculate a relative customer value over a maximum of 90 days (for the time being).

Using this cross-referenced data, Google will generate different audiences which will be imported into your Google Ads account and which you will be able to remarket using different bidding strategies (we generally recommend a Target CPA or Target ROAS strategy).

It seems that this ratio is really effective from 1000 conversions/week upwards.

New audiences to come

It has to be said that some accounts have audiences with little or no data.

For example, the Google Ads audience for household income:

The Google Ads teams have announced that major progress is going to be made on these reports in order to provide advertisers with as much data as possible for their optimisation.

The new audiences:

• Parental status : currently limited to a simple yes/no, the user's parental status will be enriched with data concerning children: number, age and level of education.
• Marital status : this new audience will show whether the user is single, married or in a relationship (outside marriage).
• Level of education: the user's level of education - Bac and Master.
• Housing: Does the user own or rent their home?

See all software Search Engine Marketing (SEM)

The importance of audience strategies in a context of automation

Google Ads is moving towards greater automation of campaign management (Smart Bidding, Smart Shopping, etc.). To remain in control of their actions and their machines, advertisers will have less need to be technicians, but will need to strengthen their strategic approach.

Audience strategies are therefore essential to master in this context: they will make it possible to optimise costs by excluding unprofitable segments and to maximise volumes on high value-added audiences.

Sponsored article. The expert contributors are authors who are independent of the appvizer editorial team. Their comments and positions are their own.