Choosing and configuring the right type of campaign with Ally

An introduction to Ally, your AB Tasty virtual assistant

Ally is your virtual assistant – a module that can guide you towards the right type of campaign for your needs. Ally will pose various questions about your objectives, the type of message or version you want to create, the segment you want to target, and so on. This will help you select the best campaign for your specific objectives. It is important to select the right type of campaign for the following reasons:

• Reports differ according to the type of campaign, as their purpose is to meet the user’s expectations. In order to create the best campaign insights, you need to choose the right type of campaign to begin with.
• Regarding campaign duplication, the AB Tasty matrix provides actions that are specific to the type of campaign you choose: for example, transforming a winning variation into a temporary patch or transforming an AB test into a multipage test. Therefore, choosing the right type of campaign beforehand will help you optimize your website over the long term.

The final screen of the Virtual Assistant displays a recommendation on the type of campaign you should create, with a definition, a tip, and a link to specific tips you can also find in this article.

Good to know 💡

From the final screen, we recommend clicking the dedicated helper to open it in the right panel (Need help?). You can also click the Create button to open the Main Information Step of your newly created campaign. This way, it will serve as a reference during the configuration process of your campaign. The tips provided in this helper will be very useful as you configure your campaign.

Tips to configure your campaign properly

In this section, we have gathered digest content, to-do lists, and tips to help you in your campaign-creation workflow.

🔑 Configuring an A/B Test

Definition

An A/B test enables you to test the performance of a new version of an element on your website (for example, a CTA, header, image, or different wording). After analyzing the results of your test, you can decide which version has performed best according to the goals you wanted to reach (increasing the number of clicks, the number of pages viewed, decreasing the bounce rate, etc.). You can then apply these changes directly to your website.

To learn more about A/B tests, see our complete A/B Test Guide. 

Configuration

Follow these eight tips to successfully set up an A/B test:

1. MAININFORMATION: We recommend establishing a hypothesis for your test using the following model: If I apply [this change on my webpage] for [this audience], then [it will impact] and enable / enhance [this goal]. Add this hypothesis in the test’s description field. In the URL field, enter the URL you want to load in the editor. This URL may well be the one you target in the targeting step.
   
   
2. EDITOR: Make simple modifications such as changing the color of a button, replacing an image, or changing the wording of an element. To test combinations of changes simultaneously, choose the Multivariate test. If it’s clickable, always put action tracking on the element you modify.
   
   
3. GOALS: As the primary goal enables you to determine which variation takes precedence over others, you should choose action tracking related to the element you have modified. After all, it’s the user behavior that’s more likely to be affected by the modification you have made in the editor.
   
   
4. TARGETING: Be careful when entering the targeted pages (in the Where section) to ensure that your test will display correctly on the pages you want to target (e.g. Product pages). To trigger the test only for visitors using a specific device, use the Device criterion in the Triggers section.
   
   
5. TRAFFIC ALLOCATION:Traffic allocation must be identical for each variation. For example, avoid uneven distributions, such as: original 20%, variation one 50%, variation two 30%.
   
   
6. QA: QA is a fundamental step to ensure that modifications appear correctly on all targeted devices and that your action tracking has been configured correctly. For more information, refer to our QA Guide.

1. Before launching your test (making it visible to your audience), make sure the QA mode is disabled, the targeting of your test has been saved, and traffic allocation is evenly divided between variations.
   
   
2. We recommend you run a test for at least 15 days before analyzing the results. For more information on why, refer to the readiness section in the Reporting Guide. 

Use cases – A/B tests can be used in the following cases: 

🔑 Configuring a Multipage Test

Definition

A Multipage Test enables you to try out a new version of one or several elements across a user’s journey, including on different pages that don’t share the same structure (such as the homepage, the product pages and the basket page). These elements may have different layouts depending on page structure.

As with A/B tests, after analyzing the results of your test, you need to decide which version has performed best according to the goal you wanted to reach (increasing the number of clicks, the number of pages viewed, decreasing the bounce rate, etc.). You can then apply these changes directly to your website.

To learn more about Multipages Tests, refer to our complete Multipage Test Guide. 

Configuration

Follow these tips to successfully set up a Multipage Test:

1. MAIN INFORMATION: We recommend establishing a hypothesis for your test using the following model: If I apply [this change on my webpage] for [this audience], then [it will impact] and enable / enhance [this goal]. Add this hypothesis in the test’s description field. In the Pages section, enter the URL you want to load in the editor for each page or group of pages. Each page coincides with a part of the user's journey you want to test. You need to include at least two pages that must be different (e.g.: the product pages and the basket page).
   
   
2. EDITOR: You must adapt the modification to each page you have configured. Each page has the same number of variations, as you build an entire new user’s journey.
   If it is clickable, always put action tracking on the element you add or modify in order to track the performance of your campaign (for example, a cross to close the pop-in, a CTA in the pop-in, etc.). Don’t forget to put action tracking on each page.
   
   
3. GOALS: As the primary goal enables you to determine which variation takes precedence over the others, for Multipage Tests, exceptionally, your primary goal should be the result of the user’s journey, regarding the main goal of your test: retention (the primary goal should be the number of viewed pages), loyalty (revisit rate), or conversion (transaction rate). You should also choose action tracking related to the elements you have modified as secondary goals. After all, this is the user behavior that’s most likely to be affected by the modifications you have made in the editor.
   
   
4. TARGETING: The target pages (the Where section) must be different for each page, as they relate to a specific step in the user’s journey. The segment (the Who section) must be the same for each page. Don’t forget to configure targeting for each page. If one or several sections (Who, Where, How) have the same configuration, you can use the Replicate targeting option.
   
   
5. TRAFFIC ALLOCATION:Traffic allocation must be identical for each variation. For example, avoid uneven distributions, such as: original 20%, variation one 50%, variation two 30%.
   
   
6. QA: QA is a fundamental step to ensure that modifications appear correctly on all targeted devices and that your action tracking has been configured correctly. For more information, refer to our QA Guide.

1. Before launching your test (making it visible to your audience), make sure the QA mode is disabled, the targeting of your test has been saved, and traffic allocation is evenly divided between variations.
   
   
2. We recommend you run a test for at least 15 days before analyzing the results. For more information on why, refer to the readiness section in the Reporting Guide. 

Use cases –Multipage Tests can be used in the following cases: 

🔑 Configuring a Multivariate Test

Definition

A Multivariate Test enables you to test combinations of changes simultaneously on your website. You can change several elements on a page at the same time (for example, the color and wording of a CTA) and identify which of the possible combinations performs best according to the goal you wanted to reach (increasing the number of clicks, the number of pages viewed, decreasing the bounce rate, etc.). You can then apply these changes directly to your website.

Unlike an A/B test, which involves testing each hypothesis separately, a Multivariate Test allows you to run hypotheses at the same time and to find the best combination of all variations. The purpose of Multivariate Tests is to measure the interactive effects between several supposedly independent elements (e.g. page title and visual illustration).

A subtest contains variations that will be tested independently across multiple combinations. For example, if your goal is to test a simple CTA, you might want to create three subtests: Color, Shape, and Wording, with as many variations of colors, shapes and wording as necessary. AB Tasty will create all the possible combinations and display them randomly to your traffic. 

To learn more about Multivariate tests, see our complete Multivariate Test Guide. 

Good to know 💡

The number of combinations created is calculated as follows: [Number of variations of the first subtest]*[Number of subtests]. The more subtests with variations you configure, the more traffic you need to allocate to the different combinations. In this case, it may take longer to obtain accurate data and results from your campaign than a classic A/B test. 

Configuration

Follow these nine tips to successfully set-up a Multivariate Test:

1. MAIN INFORMATION: We recommend establishing a hypothesis for your test using the following model: If I apply [these changes on my webpage] for [this audience], then [it will impact] and enable / enhance [this goal]. Add this hypothesis to the test’s description field. By default, only one subtest is created, but you can add as many subtests as necessary. For each subtest, enter the URL you want to load in the editor. In most cases, you must enter the same URL for all subtests.
   
   
2. EDITOR: For each subtest, you can add as many variations as you want. For example, if your first subtest is related to the color of a button, you can test the color blue in variation one and red in variation two.
   
   
3. GOALS: You can’t select any goals for a Multivariate Test – only the transaction goal is available and preselected. This is also why, in the editor, you can’t add action tracking or tracking widgets.
   
   
4. TARGETING: Be careful when entering the targeted pages of your test (in the Where section) to ensure your test will display correctly on all the pages you want to target (e.g. product pages). To trigger the test only for visitors using a specific device, use the Device criterion in the Triggers section. If the Who, How and Where sections have the same configuration for multiple subtests, you can use the Replicate targeting option.
   
   
5. TRAFFICALLOCATION: All users must be tracked. Each percentage of the targeted traffic is assigned to a subtest.
   
   
6. QA: QA is a fundamental step to ensure that modifications appear correctly on all targeted devices and that your action tracking has been configured correctly. For more information, refer to our QA Guide.
   
   
7. Before launching your test (making it visible to your audience), make sure the QA mode is disabled and the targeting of your test has been saved.
   
   
8. We recommend that you let your Multivariate Test run for at least 15 days before analyzing the results, and that you wait until at least 5,000 unique visitors have been tested per variation and per combination (with around 500 conversions). This will help to guarantee reliable results. For more information, refer to the readiness section in the Reporting Guide.
   
   
9. REPORT: Your report displays a breakdown of the “weight” of each variation within the combination. The weight of a variation is the impact a particular variation has on the improvement identified in this combination.
   
   

Use cases –Multivariate Tests can be used in the following cases:

🔑 Configuring a Split Test

Definition

A Split Test or Redirect Test enables you to test the performance of a new page, created and hosted outside of AB Tasty. This page is used as a variation within an A/B test campaign.

This type of test is useful when modifications involve a significant amount of work (for example, a brand-new design).

In this case, two URLs are involved:

• The source URL: the one corresponding to the original version. This is the URL from which you want your visitors to be redirected.
• The destination URL: the new page used as a variation. This is the page visitors will be redirected to when they land on the source URL.

As in a classic A/B test, after analyzing the results of your test, you will see if the destination URL has performed better than your original version depending on the goal you wanted to reach (for example, increasing the number of clicks, the number of pages viewed, or decreasing the bounce rate). 

To learn more about Split Tests, see our complete Split Test Guide.

Configuration

To create a Split Test or Redirect Test, first select A/B test as your campaign type, as the Split Test is not a campaign type in and of itself.

Follow these tips to successfully set-up a Split Test:

1. MAIN INFORMATION: We recommend establishing a hypothesis for your test using the following model: If I redirect [this URL to this new one] for [this audience], then [it will impact] and enable/enhance [this goal]. Add your hypothesis in the test’s description field. In the URL field, enter the URL you want to load in the editor. This URL should be the source URL. This is important if you want to set up tracking on this page, including action tracking.
2. 
   EDITOR: Variation one corresponds to the redirect page. To create the redirection, click on Variation one in the header navigation and select the redirect option. You can then paste the URL of the page you want to redirect to in the pop-in. We recommend ticking the following boxes:

• • Check the accessibility of the page – This option automatically pauses the test when the redirected URL is no longer online (with an email alert to the admins of the account). Once your landing page is available again, another email will be sent and you will be able to restart the test again.
  • Keep AT internet referrer in URL – If you use AT internet, this will ensure you keep the original URL.
  • Enable redirection – This option allows AB Tasty to perform the redirection. You must select it.
  • If you want to redirect to several pages, you can select the Regular expression option and enter the regular expression in the URL field.

Heads up ⚡In a redirected variation, you cannot add Variation JavaScript or make changes. Turning an existing variation (with changes) into a redirection variation deletes all changes. Only action tracking can be added, as it is not specific to a single variation but to the whole campaign.

1. GOALS: As primary and/or secondary goals, you can choose action tracking that are common to the original variation and to the new page. For more information, see the relevant section of our Goal Configuration Guide.
   
   
2. TARGETING: You must target the source URL.
   
   
3. TRAFFICALLOCATION: Traffic allocation must be identical for each variation.
   
   
4. QA: QA is a fundamental step to ensure that the redirect page appears correctly on all targeted devices and that your action tracking has been configured correctly. For more information, refer to our QA Guide.
   

Before launching your test (making it visible to your audience), make sure QA mode is disabled, the targeting of your test has been saved, and traffic allocation is evenly divided between variations. Make sure you have enabled redirection in the editor step.

We recommend you let a test run for at least 15 days before analyzing it. For more information, refer to the readiness section in the Reporting Guide. 

Heads up⚡

If the AB Tasty tag is not implemented on the page that you want your visitors to be redirected to, they will be able to view the page but no data will appear in your campaign reporting.

🔑 Configuring a Simple Personalization

Definition

A Simple Personalization enables you to personalize the experience of one specific segment of visitors on one page (or another page of the same type) of your website.

Good to know 💡

A personalization campaign is always based on an association between a specific message and a target. Therefore, it is important to identify and create your visitor segments beforehand. These will depend on your industry. You will then be able to build your personalization strategy based on these segments and on audience KPIs. For more information, refer to our Segments Guide.

To learn more about Simple Personalizations, see our complete guide. 

Configuration

Follow these nine tips to successfully set up a Simple Personalization:

1. MAIN INFORMATION: We recommend establishing a hypothesis for your personalization campaign using the following model: If I personalize [this element on my webpage] for [this audience segment], then it will impact [this behavior] and enable / enhance [this goal]. Add this hypothesis to the description field of your campaign. In the Page section, enter the URL you want to load in the editor.
   
   
2. EDITOR: You must create your message in the unique loaded view. Add JavaScript and / or CSS code, use the visual editor, or add a widget from the library to create your personalized message. We recommend adding action tracking to the element you add or modify in order to track the performance of your campaign (for example, a cross to close the pop-in, a CTA in the pop-in etc.).
   
   
3. GOALS: Choose the action tracking(s) you have configured in the editor as primary and secondary goals. You can also add specific goals such as transaction (for e-merchants and conversation-based websites) or the number of viewed pages (for media and engagement-based websites).
   
   
4. TARGETING – the WHOsection: This is the most important section, where you should choose the right segment as a function of the message you have created in the editor.
   
   
5. TARGETING – the WHEREsection: Choose the unique URL (or a saved page) or the type of pages with the same layout (such as product pages) on which your message will be visible. The URL used in the editor step remains a sample of the URL(s) you configure in this step.
   
   
6. TARGETING – the TRIGGERsection: This step is optional. You can add specific session-based triggers, such as a required number of viewed pages before displaying a message, the landing page of the session, and so on.
   
   
7. TRAFFICALLOCATION: If you allocate 100% of your traffic to your experience, you won’t be able to make a comparison with the original version. We recommend directing 20% to 30% of your traffic to the original version (depending on the amount of traffic on your website) to be able to track successfully the performance of your personalization.
   
   
8. QA: QA is a fundamental step to ensure modifications appear correctly on all targeted devices and that your action tracking has been configured correctly. For more information, refer to our QA Guide.
   
   
9. Before launching your campaign (making it visible to your audience), make sure the QA mode is disabled. 

Use cases – Simple Personalizations can be used in the following cases: 

🔑 Configuring a Multipage Personalization

Definition

A Multipage Personalization enables you to personalize the experience of one specific segment of visitors across their full customer journey, including on different pages that don’t share the same structure (such as the homepage, the product pages, and the cart page).

Good to know 💡

A personalization campaign is always based on an association between a specific message and a target. Therefore, it is important to identify and create your visitor segments beforehand. These will depend on your industry. You will then be able to build your personalization strategy based on these segments and on audience KPIs. For more information, refer to our Segments Guide.

Or to learn more about Multipage Personalizations, see our complete guide. 

Configuration

Follow these five tips to successfully set up a Multipage Personalization:

1. MAIN INFORMATION: We recommend establishing a hypothesis for your personalization campaign using the following model: If I personalize [these elements on this / these page(s)] for [this segment], then it will impact [this behavior] and enable / enhance [this goal]. Add this hypothesis to the description field of your test. In the Pages section, enter the URLs you want to load in the editor. You need to include at least two pages, as you are creating a cross-page experience for your targeted segment. For each page (different pages with different layouts), you will create a different message in the editor. These URLs are samples used to load the editor. You will be able to configure your target pages in the targeting section.
   
   
2. EDITOR: You must create a message for each page (by switching page in the header of the editor). Add JavaScript and / or CSS code, use the visual editor, or add a widget from the library to create your personalized message. If it’s clickable, always put action tracking on the element you add or modify in order to track the performance of your campaign (for example, cross to close a pop-in, CTA in the pop-in, etc.). Don’t forget to put action tracking on each page.
   
   
3. GOALS: Choose the action tracking you have configured in the editor as primary and secondary goals. You can also add specific goals such as transaction (for e-merchants and conversation-based websites) or the number of viewed pages (for media and engagement-based websites).
   
   
4. TARGETING – the WHOsection: This is one of the most important sections of your set-up, where you should choose the right segment as a function of the message you have created in the editor.
   
   Heads up ⚡
   As you are personalizing the experience of one segment on several pages, your segment should be the same for all the pages you have configured; you can use the Replicate targeting option to easily reuse the same segment across all pages.
   
5. TARGETING – the WHEREsection: You should include the URL for each page of your campaign where your message will be visible. The URLs used for the editor step remain samples of the page types you configure in this step. You can switch between pages using the dropdown menu.
   
   Good to know 💡
   A visitor doesn’t need to see pages in a specific order. For example, they can enter the website from a product page and see page two of the personalization campaign (for example, a coupon code displayed in a banner). Then, they can go to the homepage and see page one (for example, a coupon code displayed in a pop-in).
6. TARGETING – the TRIGGERsection: This step is optional. You can add specific session-based triggers, such as a required number of viewed pages before displaying a message, the landing page of the session, and so on.
   
   
7. TRAFFICALLOCATION: We recommend directing 20% to 30% of your traffic to the original version (depending on the amount of traffic on your website) to be able to measure all increments accurately (in terms of transaction rate, for example).
   
   
8. QA: QA is a fundamental step to ensure modifications appear correctly on all targeted devices and that your action tracking has been configured correctly. For more information, refer to our QA Guide.
   
   
9. Before launching your campaign (making it visible to your audience), make sure QA mode is disabled and all targetings (Where, How) have the same configuration.
   
   

Use cases – Multipage Personalizations can be used in the following cases: 

🔑 Configuring a Multi-experience Personalization

Definition

A Multi-experience Personalization enables you to personalize the experience of several segments of visitors on one page of your website. You can set up a dedicated experience for each segment, meaning that each segment will see a different message or piece of content.

This type of campaign lets you prioritize the different experiences you have created:

• If a visitor matches multiple segments (overlaps), they will see only the experience with the highest priority (as you’ve set it up in the Main Information Step).
• If a visitor matches only one segment, they will see the experience targeted to them.
• If a visitor does not match any of the segments, they won’t be targeted and won’t see the campaign.

To learn more about Multi-experience Personalizations, see this complete guide. 

Configuration

First, ask yourself if your campaign will include overlaps. For example, let’s say you plan to create two sales pop-ins on your homepage: one to display a coupon code to your VIP visitors, and one to display a coupon code for visitors who have purchased within the past month. You need to establish which pop-in is displayed to a visitor who is both a VIP and has purchased within the past month.

The most important message is assigned “Priority 1”, so visitors matching both segments (“VIP” and “Last month purchasers”) will only see the Priority 1 pop-in.

Follow these tips to successfully set up a Multi-experience Personalization:

1. MAIN INFORMATION: We recommend establishing a hypothesis for your personalization campaign using the following model: If I apply [these changes on my webpage] for [these audience segments], then it will impact [this behavior] and enable/enhance [this goal]. Add this hypothesis to the description field of your campaign. Choose the right priority for the right message: create each experience based on a specific priority. The most important experience, that is to say, the experience that will be seen first, appears at the top (Priority 1).
   
   
2. EDITOR: Each experience must display a specific message to a specific segment.
   If it is clickable, always put action tracking on the element you add in order to track the performance of your campaign (for example, a cross to close a pop-in, a CTA in the pop-in, etc.). Don’t forget to put action tracking on each experience (for each pop-in, for example, each with a different name: “Click cross pop-in VIP” and “Click cross pop-in NYC”, etc.).
   
   
3. GOALS: Choose the action tracking you have configured in the editor as primary and secondary goals. You can also add specific goals such as transaction, if your message has a potential impact on purchases, or access to a specific page of your website if your message has a pageview goal.
   
   
4. TARGETING: The segment and/or the trigger must be different for each experience. In our example, you can create a “VIP segment” for the Priority 1 experience and a “Past month purchasers segment” for the Priority 2 experience. Target pages (the Where section) must be the same for all experiences.
   
   
5. TRAFFICALLOCATION: If you allocate 100% of your traffic to the experiences, you won’t be able to track the increment of each experience (in terms of purchases for example). We recommend directing 10% to 15% of your traffic to the original version (depending on the amount of traffic on your website). Don’t forget to set up traffic allocation for each experience of your campaign (in our example, for both experiences).
   
   
6. QA: QA is a fundamental step to ensure that modifications appear correctly on all targeted devices and that your action tracking has been configured correctly. For more information, refer to our QA Guide.
   
   
7. Before launching your campaign (making it visible to your audience), make sure QA mode is disabled.

Use cases – Multi-experience Personalizations can be used in the following cases: 

🔑 Configuring a Patch

Definition

A patch enables you to temporarily correct an element on your website, either to fix an issue or to bypass a long release process for a UX improvement. A patch lets you quickly force a fix (remove a functionality, change the wording, update incorrect information) and ensure its visibility to all your traffic.

It also enables you to rapidly push a winning variation (from an A/B test) into production without the help of your technical team. In this way, developers do not need to rush out new functionality but you can still employ a winning tweak or variation almost immediately. When the developers’ full fix goes live, you can pause your patch campaign and archive it.

Good to know 💡

A patch is not a personalization, for the following reasons: 

• No data is collected: No reporting, no goals, no tracking. The only information collected is the number of unique visitors who have seen the patch (displayed in the dashboard). 
• There is no specific segment: A patch is displayed to all your visitors and all your traffic, though you can use the trigger library if needed (to display the patch on mobile devices only, for instance).

To learn more about patches, see our complete guide. 

Heads up ⚡

We recommend letting your patch campaign run for the shortest time possible. Indeed, the fix you created is only running through AB Tasty and can become outdated. When you have the available resources, you should hardcode the modifications of your patch directly on your website. 

Configuration

Patch campaigns are accessible from the test dashboard via the Create dropdown menu. Here are some tips to successfully set up a patch:

1. MAIN INFORMATION: We recommend adding a description to your patch to add useful information you want to share with your team. For example: “This patch will fix a typo error on product #447 158.” In the URL field, enter the URL you want to load in the editor.
   
   
2. EDITOR: Make your changes as usual via the code editor (using JavaScript or CSS) or the visual editor. You can’t add tracking because patch campaigns have no reporting and are designed only to improve the performance of your site. To create commercial messages, create a personalization campaign. You can’t create variations in a patch because the aim is just to create a unique fix.
   
   
3. GOALS: This type of campaign doesn’t allow you to configure any goals.
   
   
4. TARGETING:
   Segment: You can’t choose a specific visitor segment; your campaign will be visible to all your visitors.
   Page(s): Define the page(s) you want to patch by selecting the right saved page(s), custom page(s), custom code, and ID/class / element criteria.
   Trigger: This section is optional but useful to display your campaign according to specific conditions (for instance a specific device or region, or a number of pages viewed).
   
   
5. TRAFFIC ALLOCATION: This step is not available for patch campaigns, meaning 100% of your traffic will see your patch.
   
   
6. QA: QA is a fundamental step to ensure the correction(s) display properly. For more information, refer to our QA Guide.

Heads up ⚡

In AB Tasty, you can display campaigns even if your visitors have not yet given their consent. For patch campaigns, no data is collected but, by default, they are displayed only when the visitor has given their consent. To display patches to all visitors including those who have not yet given consent, go to Settings > Cookies > Cookie deposit and check the Patch box. For more information, refer to this guide on Managing your visitors’ Privacy. 

Use cases – Patches can be used in the following cases: 

Use case 1: Creating a patch from scratch

You can create a patch from scratch, for example: 

• To correct a typographical error
• To modify a legal mention
• To add a health message
• To hide a CTA (for example, when experiencing a stock shortage)

Use case 2: Transforming a winning test variation into a patch

To transform a winning test variation into a patch, follow these steps:

1. From the Dashboard, hover over the test you want to use as a reference for your patch.
2. Click, then select Duplicate.
3. Choose Patch from the first dropdown menu.
4. Choose the variation you want to push to 100% of your users (the winning one) and validate. Your patch appears in the test dashboard.

Your campaign set-up is imported from the source test but you can add or modify information if needed. Make sure you have included a description of your patch campaign.

Good to know 💡

You can also duplicate the winning variation of a Multipage Test into a patch. In this case, you must create one patch per page as this type of campaign does not allow you to have several pages.