Updated 2/12/25.
LLMs like ChatGPT Deliver Profit Now
Imagine hiring dedicated professionals for each area of web marketing, who all grasp ~95% of their profession, comprehend 100% of your instructions and your firm’s defining documents, produce at 10x human time, all for $20/mo. That’s AI/LLM in a nutshell.
LLMs (Large Language Models, like ChatGPT) may be over-priced in the stock market, but a quarter of that valuation is still huge, and that value reflects firms actually buying and using the tech now. The question is, how should your firm use LLM tech? DISC’s 28 years working with search engines, which essentially are LLMs, moved DISC into expertise in answering this question and moving clients to profitable adoption.
It happens in two stages:
Those two parts each have 3 Tiers that stand alone. DISC will provide each Tier’s details in your proposal. Each Tier includes the prior, so starting with Tier 1 and later doing Tier 2 adds only the difference in cost.
Customized LLM “GPTs” are ideal for content-level SEO and CRO (conversion rate optimization). By using it, you learn essentials of LLM application to other areas of business (and life).
1) Optimize Your Cloud for AI/LLMs (OCA)
Embracing AI/LLM accelerates profit now, but most companies aren’t prepared. Proper setup and simple guidelines reduce LLM inaccuracies and hallucinations while maximizing “token limits,” so that LLMs accomplish what you want.
Proper setup begins with Optimizing your Cloud (file system) for AI/LLMs (OCA, pronounce “okay”). DISC ensures you integrate both long-established and AI-specific practices in your firm’s file and knowledge systems. This enables an LLM assistant to amplify and expedite human expertise to achieve marketing and business goals.
DISC’s OCA Process Includes:
- A questionnaire.
- Audit of your cloud’s LLM readiness.
- A final report with tasks prioritized by likely ROI.
Implementing the report’s tasks is mostly a clerical tech process. For implementing, the report ascertains the best coordination among your team or vendor and DISC’s IT resources. Time/cost estimates multiplied by ROI impact determines priority 1-5 of each task.
Optimizing your cloud is justified by LLM readiness alone, yet the resulting knowledge usability often delivers positive ROI regardless of LLMs. A cloud organized in line with your current workflow, one that contains all our team needs and links to subordinate knowledge only when relevant, is the very definition of the successful firm. Your people will love the new ease and efficieny.
OCA Pricing
Every firm has different quantities and usage of cloud documents, so Tier 1 Discovery is required to decide subsequent tiers.
- Tier 1, Discovery: Most DISC services begin with a low-cost Discovery phase that projects your ROI before investing more. Your final report gives valuable, actionable guidance even if ROI projections show Tiers 2 and 3 not worthwhile, which is rare. $900
- Tier 2: Audit and priority 1 through 3 of 5 cloud improvements. $4500
- Tier 3: Special and advanced audit and improvements, for complicated or unique cloud environments. $9000+, custom quoted.
Your team’s time:
- Baseline: ~3 hours to answer a questionnaire, provide cloud view-only access (small, representative samples alone will do), and to digest the final report.
- Additional time: 3 more hours each for Tiers 2 and 3.
2) Deploy Your LLM Professional “GPT”
Overview
This section outlines building your LLM SEO pro assistant (“GPT”). The process is similar for any business GPT.
OpenAI’s LLM models lead the pack most of the time, but configuring any top LLM is similar. DISC will setup your preferred LLM.
Building your custom LLM is akin to orienting a new professional employee: provide key company documents, specify objectives, and outline procedures, output formats, and success metrics.
DISC begins your LLM deployment with website SEO because this area is especially well-suited for LLM application, and most businesses seek more prospects and higher conversion rates to customers. The process of making and using an SEO GPT will train your team to use LLMs for other vital marketing and operations.
Your GPT is Pre-trained
Like a professional employee, the LLM comes pre-trained. ChatGPT, for instance, has a vast database of public and curated content, such as national, state, and local laws, Google’s ~3000 pages of SEO guidance, paywalled content from major newspapers, public domain medical studies, and much, much more. The top LLMs also include your website’s public content as of the most recent pretraining round. The P in ChatGPT refers to “pretrained,” just as, you hope, a professional employee is pretrained in his field.
Informing Your GPT
The process is straightforward: you inform the GPT of what you want, and provide relevant documents. The art and science of using LLMs is in understanding how they work and in tuning instructions accordingly. LLM failures, whether ridiculous or dangerous, almost always stem from poor setup and “prompting.”
- Your GPT’s Knowledge Base will include key proprietary documents that guide the GPT in achieving your goals. These documents may encompass product or service descriptions, target market data, web pages, and brand briefs about your firm’s mission and personality.
- GPTs’ Custom Instructions are crucial. They explain what you want the GPT to do and require the GPT explain its steps and reasoning.
- DISC provides prompting guidelines and templates to ensure optimal outputs.
Trial Run, Tuning, Finale
Your SEO GPT applies to website content. Technical SEO (server and code foundations) will likely remain beyond the capabilities of LLMs until 2028 because technical/CMS-SEO is highly contextual, precisely and dependently sequenced, and multidisciplinary.
- For a deeper explanation of this LLM limitation, see this blog paragraph under “A 2-Dimensional Model for Understanding the Integration Imperative.”) To understand what complete SEO is today, see DISC’s SEO page and the blog post, Now Core SEO Can Long Endure.
DISC tests the GPT, then monitors your team’s first use, which may entail final tuning.
As a DISC client, you get access to a GPT that covers typical Q&A about using the GPT. (The drive to automate such customer service fueled the rise of the entire LLM industry in the first place.) DISC’s 3 Tiers include follow-up consulting to ensure efficacy.
Prices
- Tier 1, Discovery: Most DISC services begin with a low-cost Discovery phase, to be sure your firm will benefit before going further. Your delivered report projects your ROI and gives valuable, actionable guidance even if it determines that proceeding further is not worthwhile, which is rare. $900
- Tier 2, SEO Writing Assistant: This is trained by your existing keywords, website content, marketing data, and brand studies. $3000
- Tier 3, Entity Structuring: DISC’s use of specialized Entity research tools trains your GPT to improve your site’s usability and conversion rates by ensuring a sound Entity structure and flow in your website. The Entity research produces Entity-clustered lists of keywords, which we feed into your custom GPT so that SEO’d and CRO’d writing adheres to the enduring Entity structure of your business and market. A graphic designer can also use this trained GPT to generate images and layout ideas that are on brand. $9000
Specialized SEO GPTs: Discovery may reveal that one or more GPTs dedicated to SEO sub-tasks would deliver positive ROI within one year, in which case DISC will quote those GPTs. Examples:
- Local and Reviews SEO (LR-SEO)
- Video SEO
- News and article site SEO
- Advanced Schema.org semantic markup.
- These GPTs are custom priced, but expect between $1500 and $4000.
Your team’s time:
- Baseline: ~5 hours to answer a questionnaire, work with DISC to setup (train) your first custom GPT “employee,” and achieve facility with using it.
- Additional time: 2 more hours per each of Tiers 2 and 3.
- Each additional GPT or LLM “employee” for whatever work you need done can entail a wide range of time investments, but each one greatly reduces the learning curve for the next.