Knowing what to do can be one of the biggest challenges for new coders. In a previous article we covered choosing a language. Now you’ve made your way around tutorials on variables, loops, and functions but how can you apply it to your job?
Many coding tutorials are geared towards full-time coding positions or domains that have nothing to do with marketing. This leaves learners with the burden between learning and implementing.
This post will cover a few example and starter ideas. We won’t completely remove the burden but we will at least approach is from a better angle.
The Goal of this Post
The goal is not to walk you through a complete solution but to provide broader ideas on where to start. You can then take these ideas and see how they apply to your own accounts.
Once you get started you will need to do some investigating of your own to piece together the steps for your solution. Don’t consider that a roadblock as much as an opportunity to work on your search skills. Searching for examples and solutions to similar tasks is a huge part of coding!
Reporting can be extremely time consuming. This is especially true to accounts that segment data in many ways (location, product line, etc) or when you have many accounts to report on (such as an MCC).
You can start from scratch and start analyzing metrics but that process can be unfocused if you are still learning your way around the language.
I recommend replicating a report you’ve already created or currently use. It could be a spreadsheet, a Data Studio report, anything that gives you some guidance on what to build. This ensures that you spend more time solving for a problem rather than figuring out what the problem is.
If you aren’t sure where to start consider building,
- A budget tracking tool
- A script to identify top and bottom performers
- Compare two date ranges
- Quality score analysis tool
- Audience/Geo analysis tool
- A/B Testing tool to identify ads to remove from rotation
Web scraping uses your script to pull the HTML from a chosen site. It is a big topic on its own but there are several helpful libraries and tools to help you along the way.
The biggest bonus to this project is learning more about how a website is built and how to identify and isolate pieces of the site. As a marketer this isn’t a huge part of your day to day but can make tagging much easier. It also gives you the language to better communicate site issues with web teams.
You can build a tool to,
- Check HTTP response codes
- Pull all business locations
- Collect product information such as makes and models (helpful for building campaigns)
- Check landing pages for key terms or language
If you pursue this check out the resources below to start you down the path.
Beautiful Soup for Python
rvest for R
UrlFetchApp for Google Scripts
This is a grab bag topic. The previous two sections had leeway on how you implemented them, utilities can be anything that makes your life easier. A utility takes the labor out of the mundane and thoughtless work. It may also help you get around clunky spreadsheets or workbooks overburdened with formulas.
A few recent examples we’ve used internally,
- Identify keywords blocked by negatives
- Check URL parameters for every ad
- Create campaigns based on seed keywords and create campaign settings for upload
- Custom graph creator with company formatting
Now that you’ve made a choice
Once you’ve made a choice, the fun part begins. Start by identifying the smaller pieces of your goal, what steps do you need the script to do? Often times the smaller steps the better! This allows you to build your script piece by piece rather than going form nothing to solution.
Learning to break problems down will help you find answers to your problems as well. Someone else has probably run across the same issue in a different context. “How to identify poor performing keywords” is much more restrictive than “how to filter rows based on value.”
Making that transition allows you to better search for resources and identify solutions. If you get stuck always remember, what is my input and what is my desired output.