Developing marketing campaigns<\/a> with LLMs (whether directly or with agents) comes with a lot of time spent reworking prompts and often clarifying data.\u00a0 Prompts are reusable, so the time spent is worth it.\u00a0 Marketing documentation often goes stale and needs to be updated frequently.\u00a0 This is particularly true if the agent is using live docs.<\/p>\n If you are using, or plan to use<\/a>, tools that utilize knowledge bases (such as custom GPTs and many agents), they would likely benefit from some updates to their structure and naming conventions.\u00a0 Many of these things are going to be beneficial to both people and machines.<\/p>\n Quick aside: keep up with the AI news with our B2B reads.<\/a><\/p>\n Let\u2019s face it, marketing strategy docs often wind up on shelves.\u00a0 LLMs, however, use them all the time.\u00a0 If you want agents as part of your team, it makes sense to tweak your documentation to work for them.<\/p>\n The goal here is to reduce the amount of rework and adjustment needed to get consistently good results from LLMs.\u00a0 This requires the removal of some space for creativity, so if you want the variance then adjust my approach.\u00a0 I think the next 3 points are relevant to all.\u00a0 Lots of this can be solved with proper prompting so we can start with that:<\/p>\n If you are reusing prompts, which you should, this is going to save a lot of time.\u00a0 Use the relevant LLM to develop a prompt for itself using best practices and based on the documentation and goals you provided. \u00a0Open another chat and upload your docs with the prompt and see what comes out.\u00a0 Then you can iterate as needed.<\/p>\n In the end, the prompt should produce good and reasonably consistent results, though there will always be variance.<\/p>\n Just make sure that if you update\/change how you do those documents that you revisit the prompt.<\/p>\n Prompts are the top priority, but you can also adjust some more consistent elements of your documentation to further improve accuracy and reduce reworks.<\/p>\n Here are some straightforward things that I am playing around with that help.<\/p>\n Most folks we work with do this already but it is still important.\u00a0 If you have any live access, it is crucial to manage how the documents are named and organized.\u00a0 At the very least, it helps the LLMs interpret and select documents faster.<\/p>\n The titles get longer on the screen, but having that information at a glance is helpful to everyone.<\/p>\n Along the same lines as some others here, telling the LLM what not to use is just as important.\u00a0 If you are just uploading to a database and not removing old docs, or if you have a live database, you will want to indicate versions that are no longer in use.<\/p>\n This requires occasional audits of knowledge bases to look for out-of-date things like personas, branding documents, any templates, and even data sets if you are using static ones in .csvs or spreadsheets.<\/p>\n I don\u2019t mind a bit of clutter in doc names if they are communicative, so I prefer to just put DEPRECATED in the title.\u00a0 If you prefer to have a more streamlined look, you can include this information in the doc itself.<\/p>\n This one isn\u2019t all that common in human-focused documents, but negative examples are great. \u00a0Personas make an excellent use-case.<\/p>\n Specific targeting performs better.\u00a0 Negative examples might say \u201cThis persona does not care about\u2026\u201d or \u201cTitles do not include\u2026\u201d. \u00a0\u00a0This will impact anything that the personas are used for from audience-building to optimization.<\/p>\n Humans don\u2019t need this since they will remember this sort of thing, but your LLMs are coming in fresh unless they have this context.<\/p>\n Executive summaries exist for a reason.\u00a0 They are super helpful to both humans and machines.\u00a0 Include a short, summarizing paragraph that goes over the purpose of the doc, most important 2 or 3 points, and any constraints (i.e. only for X audience, or internal use only)<\/p>\n While LLMs can do a great job summarizing, I have found indicating the most important points helps to ensure those make it into the final results.<\/p>\n Not every document needs this.\u00a0 A persona, to continue the above example, are already summarized and don\u2019t need this, but a deck of personas might.\u00a0 \u00a0Another good example would be that a messaging framework doesn\u2019t need this but a full positioning document will.<\/p>\n If you are using any reports, those also need a solid summary for both human and machine users.<\/p>\n As a bonus, if you are deprecating versions, the summary is a great place to indicate that.<\/p>\n Sadly AI cannot free us from document management\u2026 yet anyway.<\/p>\n If you want the best results from LLMs, you need to give it things to work with that your competition cannot.\u00a0 Just using the publicly available information, templates, and thinking will get you the publicly available results.\u00a0 You have to add something to the process to make it worth it.<\/p>\n For those of you just dipping your toes into agents: most agents use LLMs to interpret documents for them.\u00a0 If anything, this is more important for agent-based approaches to AI than just using ChatGPT or Claude in its UI (for example with customGPTs and their knowledge bases).<\/p>\n When I was just getting started in my career, I was building websites at a local digital agency.\u00a0 \u00a0Being a marketing major, I had to learn a bit of simple programming and database management.\u00a0 The best piece of advice I got was:<\/p>\n \u201cThink as if you were the machine\u201d.<\/p>\n If you have ever done the PB&J experiment<\/a>, you know what I mean.\u00a0 It is the same thing here. You get a bit more flexibility thanks to modern technology, but the concept remains.<\/p>\n If you want to chat about how you can integrate and maximize AI in your marketing processes, reach out<\/a>. \u00a0AI is overwhelming right now, but some simple practical things can help you find the right tools and get better results.<\/p>\n The post How to Improve Marketing Docs for LLMs<\/a> appeared first on Heinz Marketing<\/a>.<\/p>\n\n","protected":false},"excerpt":{"rendered":" By: Tom Swanson, Senior Engagement Manager, Heinz Marketing Developing marketing campaigns with LLMs (whether directly or with agents) comes with a lot of time spent reworking prompts and often clarifying data.\u00a0 Prompts are reusable, so the time spent is worth it.\u00a0 Marketing documentation often goes stale and needs to be updated frequently.\u00a0 This is particularly […]<\/p>\n","protected":false},"author":1,"featured_media":0,"comment_status":"open","ping_status":"open","sticky":false,"template":"","format":"standard","meta":{"footnotes":""},"categories":[98],"tags":[27158,12638,18306,8402],"dealstore":[],"offerexpiration":[],"class_list":["post-338885","post","type-post","status-publish","format-standard","hentry","category-marketing","tag-docs","tag-improve","tag-llms","tag-marketing"],"yoast_head":"\n![\"\"](\"https:\/\/www.heinzmarketing.com\/wp-content\/uploads\/2024\/01\/Predictable-Pipeline-Workbook-CTA6.png\")
<\/a><\/p>\nShow docs\/goals to LLM and have it draft the prompt<\/strong><\/h3>\n
Other stuff<\/strong><\/h2>\n
Get descriptive with your titles<\/strong><\/h3>\n
\n
\n
Deprecate old versions<\/strong><\/h3>\n
Include Negative Examples<\/strong><\/h3>\n
![\"\"](\"https:\/\/www.heinzmarketing.com\/wp-content\/uploads\/2025\/12\/negative-example-1.png\")
<\/div>\nSummarize (for important docs)
\n<\/strong><\/h3>\nDoc work never ends<\/strong><\/h2>\n
![\"An](\"https:\/\/www.heinzmarketing.com\/wp-content\/uploads\/2025\/12\/llm-pbj-768x768.png\")
<\/p>\n