The classification ability in Adobe Analytics will quickly become one of your most beloved features. Simply put, it is a way to assign different "buckets" or values of information on various traffic and conversion variables in Adobe Analytics. For example, a common conversion variable that any eCommerce organization would need is the Product ID. Adobe Analytics allows you to customize data so that whenever a specific Product ID is purchased on the site, it is automatically recorded in Adobe Analytics with all necessary visits, revenue, and any other necessary metrics tied to it. Now the problem, in the beginning, is, when you first start off your Product ID's are listed as just that in Adobe Analytics, Product IDs. 1123664, 28820, 212555.21, 6677633, so on and so forth, and unless your entire marketing team has the capability to memorize all the Product IDs your company has to offer, this is not data that is efficiently used. Now enter classification, in this scenario, the classification would allow marketers to import historical data so that Adobe Analytics could actually show the full names of products instead of just Product ID's, and it enables an extra layer of correlation that you can use in your report views.
That sounds great, but how do I do it?
Classification in Adobe Analytics isn't the easiest task to undertake, but we're going to provide you step by step instructions on how to do it, and even some tips to help make the process easier.
Step by step instructions for classification in Adobe Analytics
1. Log into your Adobe Analytics account
2. Navigate to Admin > Report Suites
3. Hover your mouse over "Edit Settings," and depending on what variable (traffic or conversion) you're working with will dictate what you should mouse over to pick the classifications.
4. In the select classification dropdown menu, choose the specific variable you want to create the classification segment for.
5. The screen will change and now will now show your classification type, with the designated variable. Mouse over the small arrow pointing to the right, and "Add Classification."
6. Once you have created the designated classification segments, navigate to Admin > Classification Importer. In the top navigation, choose browser export.
7. In the source section, ensure that your proper report suite & data set is chosen and click export file at the bottom.
8. The file will export as a .tab file, use your computer settings to make it open in Excel so you can modify it as needed.
9. You'll see that the classifications you created in step 6 are present here as headers, now you'll be able to populate the data accordingly.
Spreadsheet tips
Finding your way through Excel effectively is the hardest, and also most time-consuming portion of the entire classification process, especially if you're new to Excel, and even more so if you have a very large dataset to import. Unfortunately, there are not automatic ways to make Excel do what you want automatically, so I'd highly recommend brushing up on some basic Excel functionality before tackling a big classification project. I have found some invaluable tips on www.mrexcel.com, but a quick Google query with your specific question should yield some positive results. As a general rule, you'll want to follow these guidelines:
Prior to spending lots of time working the organization of the exported spreadsheet, I'd highly recommend saving a copy of it as a normal Excel file. This way you can save all of your edits and formatting, and reverting it to a different file type takes a matter of seconds once all of your data is organized as you wish.
When you're ready to import the organized file with all the populated data back into Adobe Analytics, you'll need to specifically save the file as a Windows Formatted Text (.txt) file that we can change it into a .tab file (explained later).
Don't delete the original Excel file, it can be invaluable later especially if the classification did not go as planned!
How to bring the file back into Adobe Analytics
In order to re-import this data back into Adobe Analytics to apply the data, we need to save this file as a .tab file. Most text editors can do this, the following steps are specific for a Mac environment using Sublime Text, which I highly recommend, or you can use the same steps in a Windows environment using Notepad ++. Both programs are very powerful and are free to download.
1. Open your .txt file in your note editor of choice.
2. Conduct a save as, and remove the .txt extension and replace it with .tab
3. Acknowledge any notifications it is giving you as a result of the changed data type.
4. Now in Adobe Analytics, navigate to Admin > Classification Importer > Import File
5. Once again, ensure that the proper report suite and data set is selected, upload your.tab, and select “Import File.”
6. If all steps were followed correctly, the importer should not give you any errors and now you just have to wait for your data to be populated and reflected in Adobe Analytics. Processing time can vary from close to instant, to up to 72 hours per upload and/or delete. This largely depends on how much data you are bringing into, or removing from Adobe Analytics.
What to do if your file is too large to import back into Adobe Analytics?
If the file is too large to bring back into Adobe Analytics using the classification importer, you can bring it back into Adobe Analytics using a File Transfer Protocol (FTP).
1. Click on Admin Tools > Classification Importer
2. Click Import File, then click "FTP Import"
3. Click add new, and fill in all fields accordingly. Make sure you choose the correct Data Set as well as the correct report suite.
4. Adobe Analytics will then generate a host/login/password combination for you that you can use to import the .tab file with the FTP program of your choice.
5. IMPORTANT once you upload your data file, you must upload a FIN file that indicates the file is ready to process. The FIN file is an empty file that has the same name as your data file, with a .fin filename extension. For example, if your data file is classdata1.tab then you must upload a file named classdata1.fin. This is only applicable for the FTP import process, not for standard browser upload.
6. Quick tip, the max file size per upload is 50MB. If your file size is larger than this, it will cause an error. There is no maximum total amount you can upload for classification, but if your file is significantly larger than the 50MB limit, you will need to break it up into numerous files.
Common questions around classifications
When should I classify my data?
After Adobe Analytics has had some time to capture significant amounts of raw, unclassified data. The reason for this is that you cannot "preemptively" put data into Adobe Analytics, and let it be populated later. For example, if we reference the Product ID scenario from the beginning of the article, we would not be able to upload Product IDs that have never been purchased on the site, and have not to be recorded in Adobe Analytics, this will cause an error to display during your import.
What's the ideal amount of classifications I should have for a variable?
There is really no good answer to this, as it will vary greatly from one business to another. An eCommernce platform may have 10 classification headers per variable because that is how dynamic their data is, whereas a non-profit supporting animal welfare may only need two. The question to ask yourself and your team is, "How can I make this data as usable as possible?" Once you have that figured out, the amount of corresponding classification headers you need should come naturally.
How often should I update my existing classifications?
Again, this will largely depend on the specific business. If you're capturing so high volumes of data on the website on a daily basis, then you will likely need to update the classifications numerous times a week to make sure all new data is fitting into your classifications that you have already created. As a general rule, if you're looking in your top performing categories and are seeing raw data that isn't usable, it's time to upload a new classification!
No comments:
Post a Comment
If you have any doubts or questions, please let us know.