Community

Google Analytics has a limit on the number of characters (bytes) for custom dimensions. It is set at 150 bytes which means if you attempt to pass a group of audiences from Lytics to GA where the total is over 150 characters you will receive a warning (Parameter for Custom Dimension is too long (max 150)) in GA or Google Tag Manager and the audiences will not be received in GA. You can attempt to overcome this by reducing the name lengths of your audiences or by deselecting API access while editing the audience in question. Disabling API access means the segments will not be 'public' and therefore will not be passed to GA.

Currently, our AWS S3 export operates on full segment scans. This means that each export will contain all users in the selected audiences. Supporting diffs is under consideration for a future enhancement but is not currently available.

The best way to update an authorization is to add a new authorization for the integration that is changing. Lytics Support can help with this transition.

You can verify that data is coming into Lytics by navigating to Data -> Data Streams. The default stream, unless otherwise configured will contain your web data and then you will see your other integrations in the drop down of stream names. Each stream page will show you the last time we received data. To verify data is being exported from Lytics you can always check in the tool in question, but you can also view the history of activity for the integration from the overview page. Under Data -> Integrations click on the tile for the integration in question. If you're already running the integration you will automatically be taken to the overview page which will show you a list of running imports and exports as well as the history of events for those works. If you believe that an integration is not working as expected, please reach out to Lytics Support.

If the structure and topics of your website change, you may want Lytics to ignore old content and start analyzing new content. This would ensure that your content topics and affinities on Lytics are only based on your current site content. However, there are potential side effects of purging all previous content. In doing so, you may impact historic affinities and lose valuable user profile information. For example, domain.com/cakes previously was about birthday cakes, but after restructuring, it is now about cupcakes. When new content affinities are calculated, your users will show affinity for cupcakes, even though they formerly viewed content about birthday cakes. Without any versioning, content affinities are formed based on whatever topics are currently associated with your content, regardless of what users actually saw at a previous time. If you would like to proceed with a content refresh, please contact Lytics Support for assistance.

This is most often caused by a referral visit where we receive the referral URL in an event. Lytics will enrich any event with a url field included in it. To ensure that only your domains are being enriched by Lytics, we can set up a list of allowed domains for your account. Then filter out any existing content of an incorrect domain in your account. Please contact [email protected] for assistance.

The Lytics Content Affinity Engine provides users with a taxonomy of rich tags on millions of topics. In most cases, the Lytics Content Affinity Engine provides users with all the topic modeling they need to power great personalization. In the event that you’d like to add your own taxonomy to what the Lytics Content Affinity Engine offers, you can add custom topics to the content graph and user affinities by adding a “meta” element to the web page. Here is an example: ``` <meta name="lytics:topics" content="Pants, Jeans" /> ``` After successfully updating the meta property, the custom topics will be processed by the Lytics Content Affinity Engine within a few hours.

Within 24-48 hours of installing the Lytics tag, content topics should begin populating in Lytics. If no content is showing in the Lytics interface, please ensure there are no soft paywalls on your website and that anonymous visitors can access your content. If you do have a paywall please contact Lytics Support. We can help you allow Lytics to process URLs or to get credentials to access them.

Content classification refers to the task of assigning one or more categories or topics to a document, which can be a URL, an email, ad, or other content types. The Lytics Content Affinity Engine uses a self-updating, graph-based content model that represents every topic with its hierarchical relationship to other topics.

In some cases, you may see topics in the Lytics interface that seem completely irrelevant or unrelated. One possible explanation for this is that our content enrichment process is attempting to enrich "sparse" URLs - URLs that contain very little information and text. You can block these topics so that they do not show in the future.