For those that missed the live webinar on February 10th, 2022, the following is a recording and a summary of the interview given by Jess Peck, Technical SEO & Machine Learning Engineer at Local SEO Guide and Clearscope’s co-founder Bernard Huang. They discussed SEO toolification and the best ways for agencies and SEOs to think about their internal SEO toolset.
. .
Bernard Huang- Co-founder of Clearscope
Follow Bernard on LinkedIn here
Jess Peck – Machine Learning Engineer at LSG
Follow Jess on Twitter here
Topics Discussed:
- What does an internal tool team bring to the table?
- How do internal tool teams work with external tools and vendors?
- How should you talk to developers both internally and externally?
- What are the best tools for SEOs to be familiar with?
- What are good ways to use Machine Learning in SEO?
- Key Takeaways
What does an internal tool team bring to the table?
- They can pipeline data together
- They can help use APIs more easily
- They can facilitate ease of use with vendor tools
- They can create the stuff our analysts need
Particularly if you are an agency, there are many benefits to having an internal tools team that understands coding. A wide range of tools and data exists that is impactful of SEO strategy and implementation.
The ability to have a skilled team that knows how to speed up processes and find patterns, you wouldn’t see otherwise, is invaluable. This might be very difficult for a solo SEO pulling just crawl, Clearscope, or SEMRush data.
It also frees up analysts to have a focus on their areas of core competence and maximize productivity.
How do internal tool teams work with external tools and vendors?
APIs and data are made available to developers in a different way than to those working through a tool interface. For example, APIs can crawl more data. In the case of Google’s new crawl inspection API, it has a limit of 600 responses an hour.
This limit is rather annoying but without an API you would need to manually hit the URL inspection tool in Google Search Console by hand. So, this level of coding knowledge allows an internal tools team of developers access in a manner that would otherwise be incredibly time-consuming.
They also help to be a point person for talking with external vendors and articulating the explicit needs of an analyst. This facilitates better technical communication and the ability to work through complex problems.
What’s an example of a way an internal tool team can help?
Developers can combine your data. For instance if you have access to Google Search Console and SEMRush you can use a tool to combine the data from both sources. This may help you see new insights about links and content that may have otherwise been difficult to notice separately.
How should you talk to developers both internally and externally?
Most developers are not necessarily experts in SEO and SEOs aren’t necessarily experts in development. This can create a dynamic where SEOs might be a bit skittish when talking with developers or they may not know how to frame asks in the most effective way. For this reason, it’s useful to keep in mind a few tips when communicating with developers.
Developers (Devs) are often interested in SEO; they just haven’t needed to consider it before. Devs are often curious about the process and why Google might be built the way it is, so they may ask a bunch of questions. If you get any questions or pushback from them it’s often because they are searching for the “why” behind a problem.
If you want to ask them for a new tool or feature there are two things to consider:
- How much time will this take to create vs. how much time will it save to have this tool? A good rule of thumb is that if it’s a repetitive process it’s probably worth it but if the process requires a lot of evaluation and thinking then, it may be a fruit that isn’t worth the squeeze.
- It may help to frame it in a way that has the most impact on how devs think and work. For example, “I was thinking about a tool where the input is X and the output is y–do you think you can do this?” Try to avoid being vague and be as specific as you can around what you want as a result of the tool being built including the UI/UX component for people who will be using the tool.
Developers are your friends– they aren’t scary. But, they can go down rabbit holes with large asks and problems– so guidance on the priority of an ask or the time you want to be dedicated to completing it may be a good idea.
What are the best tools for SEOs to be familiar with?
You don’t need to code to be an SEO or provide SEO services but if you did want to learn a bit more about coding there are two places to start that would be beneficial:
- JavaScript -Learning to code JavaScript (JS) can be useful for direct SEO which is used to do things like give performance suggestions. For example, understanding JS would allow you to see how it is impacting core web vitals so you make edits that will improve ranking.
- Python – Learning to code Python can be equally helpful with crunching data and connecting different vendor tools. If you can use Excel, it’s likely you can learn Python. In terms of functionality, Python is like Excel with a rocket pack.
While not essential for an individual SEO, coding can help expand your knowledge and ability to communicate problems and form effective solutions. It’s quite difficult for enterprises to break down content without coding as it’s needed to crawl, clean, and analyze the content as well as warehouse the data and display the results with good UX.
What are good ways to use Machine Learning in SEO?
There are some good and bad ways to use machine learning (ML) in SEO that include:
| Good Ways: | Bad Ways: |
| Finding Patterns | Algo Bias |
| Getting a Broader view of content | Errors |
Using ML to find patterns or automate repetitive processes are two of its most effective uses. For example, you can use K-means clustering to get a broad view of how keywords cluster or how topics are grouped together and find patterns you wouldn’t have seen. But, the model may need human interpretation to intervene and rework different elements to focus it appropriately.
Algorithms have bias and can get things wrong so building your own algorithms is one of the best ways to help avoid these problems. This way you can control the bias and have a human set of eyes on problems that arise.
Key Takeaways
- You don’t need to code, but having developers or SEOs with a knowledge of coding will help you:
- Communicate problems and asks for vendors more effectively
- Better understand and fix coding issues that can impact a site’s ranking
- Connect vendor tools to aggregate vendor data and find novel insights you wouldn’t have seen by using the tools separately
- Use machine learning to automate repetitive tasks to save time and find patterns that would go unnoticed
No Comment