Adriel AdOps | Episode 1 | Naming conventions, server-side tagging, and ChatGPT4 

Hello there! Welcome to Adriel’s AdOps newsletter, where 6,300+ marketers get some of our best insights on ad operations, digital marketing, and marketing automation every month in an easy-to-digest email format. 

Digital marketers today are still spending three to four hours collecting data on ad performance and creating reports, sucking up time and leaving less for strategy and growth.

In 2017, after successfully exiting Solidware, a software startup, we turned our attention to advertising. We went on to build Adriel: a marketing intelligence platform by digital marketers, for digital marketers. 

Our mission is to free your time up so you can spend more on high-value tasks rather than hours creating reports. This is why we are collecting some of our best  insights on ad operations, digital marketing, and marketing automation every month in an easy-to-digest email format. We hope you find it useful!

PS - did you catch us at the Austin Marketing Summit last week? We had such a blast chatting with lots of cool marketers around the Austin area. Don’t forget to connect on Linkedin!

⚡Power Up Ad Operations: 3 Tips 

From creating ads to optimizing campaigns, adops can be incredibly time-consuming and mentally-taxing. Every 2 weeks, we share tips and tricks in this section to make it just a little easier.

1) Set up consistent naming conventions across all platforms from the campaign to the ad level to save you time, gain higher control over campaigns, and analyze your data accurately.  

Different ad platforms require slightly different variables depending on the targeting and campaign structure. For instance, in Google Ads, you'd want to include “ad type” in your campaign name, while in Facebook Ads you can go with an audience variable. You can break down variables into two categories by the information they provide.

When naming your campaigns, arrange these variables from the broadest to the narrowest. For example: 
[GEO] - [BRAND] - [PRODUCT CATEGORY] - [PRODUCT]
US - Nike - Basketball Shoes - Air Jordan

2) Use server-side tagging to regain control of your data. With identifiers gone and third-party cookies soon to be extinct, it’s getting tougher to map out customer journeys and deliver personalized experiences. Server-side tagging can help. 

Instead of encrypting pixels to the web container and having HTML tags send the data to ad networks, server-side setup starts with creating a custom domain as an initial endpoint.

From there, datasets are remodeled into a format that’s compatible with the tag management system like the Google tag manager (GTM) before getting shipped out to 3rd party vendors.

This moves the workload from the user’s browser to a cloud server managed by the business, giving full control over the data log customers leave behind without running into privacy risks associated with client-side tagging.

3) Plan an entire marketing campaign with ChatGPT4. AI can be a massive timesaver—if you know the right prompts. LinkedIn content creator, Ruben Hassid, shares his 6-part “Super Prompt” formula to help you plan an entire marketing campaign in a few minutes. 

📖 What digital marketers are buzzing about

From creating ads to optimizing campaigns, adops can be incredibly time-consuming and mentally-taxing. Every two weeks, we share tips and tricks in this section to make it just a little easier.

1) The State of Omnichannel Marketing in 2023 Report. Marketers that use >3 channels for ad campaigns earn a 494% higher order rate than others that just use one! This Digiday x Amazon Ads Report highlights how over 110 brands and agencies are approaching omnichannel marketing to solve challenges with targeting, measurement, and ROI. Some key insights: 

✅ 6 out of 10 marketers believe the main goal of omnichannel marketing is to build trust/loyalty with customers.

✅ Marketers plan to invest 533% more in audio (podcasts and radio) and 1800%  😱 in linear TV in 2023.

✅ The top 3 KPIs marketers are tracking in 2023 are conversion rate, cost per click, and cost per acquisition. 

2) How to build a Hagakure structure in Google Ads [BLOG]. Who would’ve thought an 18th century samurai code of honor can lead to +115% conversion uplift on Google Ads? 

Hagakure is a much simpler way of structuring your Google Ads account, taking full advantage of machine learning, smart bidding, and dynamic search ads. It significantly cuts down the time, money, and effort required to manage Google Ads PPC campaigns. Get a 3-step framework to start implementing it on your Google Ads account in this blog post.

3) Paid social: lead generation engine or an awareness channel? Chris Walker’s Refine Labs is re-defining how marketers think of paid social, and they share a step-by-step guide to their unique approach to social media (used by 100+ B2B companies) in the Vault. This is a great place for marketers to get templates, vetted resources, and advice based on execution, rather than theory.

🔦Behind the Adriel Team

WHO: Kevin Autin Martineau
ROLE: Chief Machine Learning Engineer

What do you do at Adriel, and what are you working on right now?

I’m an ordinary backend engineer most of the time. But when we need some work done related to AI, I’m the guy to work on them. Right now, I’m using ChatGPT and other AI models to generate ad images and ad insights automatically.

How will AI change in the future?

Pretty soon, I think AI will be able to generate video ads based on prompts, not just images or texts. I’m not sure how fast the technology will develop but I would expect this to become common in the next one or two years. That’s already something that we [Adriel] are preparing for.

What are you most excited about?

I’m really excited about the recent advancements in language and image-generation AI models. There are so many possible applications that we haven’t even started to consider yet. The potential is really huge!

A significant part of marketers’ work is those standard and repetitive tasks that need to be typed out, like creating ads. The fact that they can be automated and made easier is incredibly fascinating.

When you’re not working with AI, what do you like to do?

I like to read books and study languages and philosophy. Nothing too crazy. I’m reading Dune right now!