​

Targeting

Ad

Landing Page

Content

Relevance, Segmentation, Standardized naming

Headlines: Relevance, unique offer, A/B iterative testing, {keyword}, extensions

Relevance, simplicity, offer, Keyword in the headline

Type

Interests, website/lookalike/similar audiences

Keywords: Exact match, phrase match

Types: Responsive Search Ads, App assets, video

Types: Desktop, Mobile (web)

Google / Apple (store)

Metrics

Metrics: CPC, CTR, CVR, CPA, CPS, LTV, Impression Share, Bounce rate

Speed score, Lighthouse, Mobile Friendly Test

App Store Optimization

Quality Score

QS: Expected CTR

QS: Ad Relevance

QS: LP experience

Experiments

Drafts & Experiments, A/B test

Ad Variations, A/B test

Ad Variations, A/B test

Budget

Predictive planning, Statistical significance

Bidding

Manual CPC, Maximize conversions, Maximize value

Dimensions

Geo, Demo, Lists, Devices

Conversions

Google ads pixel, GA import, manual / auto upload with gclid, microconversions

Settings

Languages, Geo, Search Partners

Targeting

Ad

Landing Page

Relevance, Segmentation, Standardized naming

Headlines: Relevance, offer, A/B iterative testing, {keyword}, extensions

Content: Relevance, simplicity, offer, keyword in the headline

Relevance: pick 5-10 keywords relevant to your product + add an intent word. Use Keyword Planner to research ideas.

• where to get pet nutrition supplements
• nursing jobs near me
• buy african ingredients
• how to get a cps case dismissed
• shop home crafted goods
• music lessons online
• referrals for code school grads

​

Segmentation: segment keywords into adgroups if you have a separate landing page or the words are different morphologically and logically (to make more relevant Ads).

​

Search terms: remember to go through search terms report in Google Ads and clean irrelevant queries. Performing this exercise is essential.

​

Negative Keywords: create a list (may be multiple lists) of negative keywords that you can find through search term report. This is an ongoing process until some time the account developes. Adding Negative keywords increases CTR.

​

Standardized naming: Example - uk_en_keyword_search_phrase

Relevance: Your ad should seem very relevant to someone who searches your keywords. Most of the people read only the headline, and you get higher CTR if they see keywords in the headline (google will highlight those keywords). Apart from the headline there should be a strong offer.

​

Offer: You may provide the user a value in the form of an offer to stand out from the competitors. This could be a discount, promo-code, any incentive or special promo. This information could be placed on the headline 2 or in the description line. It’s highly recommended to write some numbers in relation to the offer: x%, x$, for free.

​

A/B iterative testing: ads should be tested against each other (you can utilise the setting “Do not optimize: Rotate ads indefinitely” to evenly distribute impressions and compare the performance. For testing Landing Pages please refer to ad variations. Compare only 1 element per iteration.

​

Consider adding {keyword} parameter in one of the headlines. Make sure all keywords look good and make sense semantically as this function automatically places the keyword into the headline/description line (30/90 symbols)

​

Extensions: Make sure to add sitelinks with description lines, structured snippets, callouts and if applicable price extensions, to make your ad more visible and to increase CTR.

​

Remember to follow editorial guidelines.

Relevance: user is expecting to see information related to the ad and keyword used, it’s important to send the user on the right landing page (if many). Sometimes it’s worth to create more landing pages for different segment of users.

​

Simplicity: more simple it is, more CVR it will bring. Try to use html/css pages with minimalistic design and simple form or button. Techstars LP is a good example.

​

Offer: make the offer visible on the first screen if mentioned in the ad text.

​

Keyword in headline: Consider adding the popular keyword in the headline on the landing page.

Targeting

Ad

Landing Page

Interests, website/lookalike/similar audiences

Keywords: Exact match, phrase match

Types: Responsive Search Ads, App assets, video

Types: Desktop, Mobile (web)

Google / Apple (store)

Exact Match: Try to use this match type for the keywords that bring conversions, this makes the phrase you use in the keyword match exactly what the user is typing (without additional words or synonyms). Example:

Keyword - [nursing jobs near me]

User query - nursing jobs near me

​

Phrase match: respects the order of the words, but shows to additional searches. Example:

Keyword - “nursing jobs near me”

User query - nursing jobs near me los angeles; how to find nursing jobs near me

​

​

Ignore broad match for now (writing keywords without operators [like] “this”) as it will use synonyms automatically and might bring irrelevant traffic in huge volumes. This could be tested later for expansion.

Responsive Search Ad (RSA): Up to 15 headlines (30 symbols) and 4 description lines (90 symbols).

Desktop and mobile version of the site or responsiveness.

App store landing page optimization

Targeting

Ad

Landing Page

Metrics: CPC, CTR, CVR, CPA, CPS, LTV, Impression Share, Bounce rate

Additionally: Speed score, Lighthouse, Mobile Friendly Test

CPC - Cost Per Click

CTR - Click Through Rate

CVR - Conversion Rate

CPA - Cost Per Action (i.e. registration)

CPS - Cost Per Sale

LTV - LifeTime Value

Impression Share (IS) - share of impressions that your ads receive vs competition

Bounce rate - % of users dropped out without engaging

Speed Score is important for the QS (Landing Page Experience component).

​

Lighthouse - improvement suggestions for the Landing Page.

​

Mobile Friendly test - improvements for mobile page

Targeting

Ad

Landing Page

QS: Expected CTR

QS: Ad Relevance

QS: LP experience

This is the most important factor in determining QS.

​

How to improve:

​

• Use Exact/Phrase Match
• Use bid adjustments for demography, audiences, locations, devices to increase CTR (only with manual bidding)
• Raise bids slowly to show on top more frequently
• Use fewer keywords to get statistics faster
• If the keyword has low search volume mark - use wider or broader phrases

How to improve:

​

• Keyword in the ad text
• Add extensions
• Use display URL and add keywords there if applicable
• Consider adding {keyword} parameter in the headline

How to improve:

​

• Keyword in the headline
• Loading time (Core Web Vitals)
• Mobile friendliness

QS is calculated for each search term. At Keyword level tab in the interface you see a QS for exact phrase, so for a broader types of keywords this metric is not precise.

Targeting

Ad

Landing Page

Experiments: D&E

Experiments: Ad Variations

Experiments: Ad Variations, A/B test

Utilise Drafts & Experiments to run A/B tests for the following:

• Bidding strategies (manual vs auto bidding)
• Settings change
• Any other substantial change to the campaign structure

​

Drafts & Experiments interface shows statistical significance of the metrics compared between original and draft versions.

Use Ad Variations to test an element of the ad with statistical calculator built in.

​

Just like in Drafts & Experiments, interface shows statistical significance of the metrics compared between original and draft versions.

Use ad variations to get landing pages with statistical calculator built in.

​

Utilize tools like Optimize to test website hypothesis.

Targeting

Ad

Landing Page

Predictive planning, Statistical significance

Calculate the budget for the campaign in the following way:

​

1. Identify what would be your desired (profitable or breakeven) CPA / CPS
2. Estimate for the initial test you’d need ~1000 clicks per geo / keyword (to measure factual performance), let’s say we aim for 1000 clicks
3. Imagine your desired CPA is 100$.
4. You have to forecast the conversion rate (CVR) as well. It’s hard to do until you run the test, but you can check from other channels (organic/facebook/direct traffic etc). Let’s imagine, your CVR is 2%.
5. Budget = 100$ CPA * 20 conversions (2% out of 1000 clicks) = 2000$
6. Average CPC = 2000$ / 1000 clicks = 2$
7. Test the traffic for at least a week. 2000$ / 7 days = 285$ daily budget on Google Ads.

​

You never know how efficiently the campaign will run, but you have to start with some value for the Max CPC and daily budget.

​

Remember, you have to factor in statistical significance for the numbers you acquire. For example, having only 10 clicks on your keyword would not represent its potential, but 1000 clicks might do.

Targeting

Ad

Landing Page

Manual CPC, Maximize conversions, Maximize value

There are different bidding strategies, but I will focus on the main, most effective ones:

​

1. Manual CPC - use this for all the new campaigns, assign CPCs by yourself based on your budget forecast.
2. Maximize conv. - as soon as your campaign reaches at least 30 conversions in the last 30 days, you can test this autobidding strategy. Assign max CPA at historical level and decrease by 10% each week until you reach the target.

​

​

​

​

​

​

​

​

​

​

​

​

• Maximize value -same as max conversions but optimizes towards value

​

​

​

​

​

​

​

​

​

​

​

​

Targeting

Ad

Landing Page

Dimensions: Geo, Demo, Lists, Devices

Dimension are layers of targeting that you can optimise in your campaigns. This might bring your CPA down significantly.

​

1. Geo - pick the geos and create tier 1, 2, 3 campaigns if you want to split the budget. Otherwise use bid adjustments in the targeted locations for the campaign. Check user location report.

​

• Demography - apply bid adjustments for Age, Gender, Household income.

​

• Lists - you can add audience lists in the campaign. Use all site visitors, similar to converters, in-market audiences to layer on top of the campaign and use bid adjustments. Check popular in-market lists in audience manager tab.

​

• Devices - optimise bids for Desktop, Mobile and Tablet based on the performance of the device type.

Targeting

Ad

Landing Page

Google ads pixel, GA import, manual / auto upload with gclid, microconversions

1. Google Ads pixel. This is a small code snippet that will record conversions. Note that google ads pixel will count conversions differently than Google Analytics. It will record as a conversion any user who clicked on the ad, even if they converted from the other channel in the end.

​

• Google Analytics import - this will attribute conversion as a last non-direct interaction by default.

​

• Manual / auto upload with GCLID. This is the way how you can manually upload conversions. GCLID is an identifier of the click, it is assigned by Google for each click and looks like this yoururl.com?gclid=1234. You need to store this parameter from the url in your database and then update the lead when sale happened. You can upload the conversion back into the Google Ads using a simple spreadsheet.

​

• Microconversions - if you don’t have a lot of sales (<30 per month from Google Ads), consider adding medium, for example a registration, in addition to the sale. This will help to optimize campaigns quicker.

Targeting

Ad

Landing Page

Settings: Languages, Geo, Search Partners

1. Languages - start with all languages unless you have any specifics
2. Geo - include targeted locations (countries and separately large cities), check the box “People in or regularly in your targeted locations” in the Location Options
3. Search Partners - this bring additional 2-10% of lower converting traffic. Apply this later in the account development (when you want to expand).
4. Leave all the rest by default