AI makes it easy to produce more content and to produce it faster. I don’t think I need to go too deep to convince anyone about how much AI, as a tool, has created efficiencies for brands, agencies, and anyone who has a goal of producing more and doing it faster.
If you’re familiar with the “good, fast, cheap” project management triangle concept, where it is commonly noted that you can only have “two” of those three things, I think that applies to AI-generated content in a lot of situations as well. But, I don’t think any of us want to intentionally create anything that isn’t “good” and doesn’t positively impact the ultimate goals.
In AI Overview SERPs, we can experience a lot of waste even if we are visible if we’re not disciplined in our research. This process focuses specifically on AI Overviews, which is the biggest intersection today with SEO in Google, but the same discipline extends to LLM visibility in tools like ChatGPT and Perplexity (with a roadmap for a future article).
In my attempt to help not sacrifice the “good” aspect and be consistent with the “fast” and “cheap” still in play (as is the expectation when we use AI), I’m sharing a specific roadmap process my team runs in my agency for clients, step-by-step, so you can borrow from it for your own work.
Step 1: Start From The Business Objective
Before we go deep into research tools, we have to know what we’re anchoring our “good” definition to. Every potential keyword, topic, or concept must map to a business outcome. Whether that is conversion to a lead, to ecommerce revenue, or some other trackable metric that has meaning for your organization.
Visibility in a SERP or traffic itself is likely not a deep enough goal unless you have a business model or outcome that is specifically defined for brand awareness or simple pageviews (which is not the majority of companies I have worked with in my career, but those in media and serving ads as their revenue source are typically the exception).
If, for example, you’re a third-party logistics provider who is relationship-driven, you’re seeking leads from businesses who aren’t interested in the lowest cost to ship an item from point A to point B. You’re looking for a business that has complex needs, is seeking a tailored solution, and an ongoing business relationship. Visibility, engagement, and ultimate content winning for this type of business is in that content investment leading to website lead form conversions that ultimately turn into booked business by the sales team and realized revenue through the course of the relationship.
No matter what type of business you’re in, and what your deepest mappable goal is, being crystal clear with that is important and is step 1 before we get into the tools and data. If you struggle with defining the ultimate goal and how to quantify it, you’re at risk of doing a lot with content, but not having it perform to the level of return on investment realization that may be expected by stakeholders.
See also: How To Define & Report SEO KPIs That Actually Move The C-Suite
Step 2: Build The Keyword Universe And Run A Content Gap Analysis
My team uses Ahrefs as our SEO and AI search research tool (we have used other tools in the past) and that’s where examples are coming from in this workflow, but feel free to substitute in/out your current tool or platform of choice.
In Step 2, we’re building a foundation. That includes creating the keyword universe that we can then run content gap analysis against.
Here are the Ahrefs tasks for this step:
- Open Competitive Analysis > Content Gap from the top nav.
- Enter your domain in the target field and add at least two to four competitors (you can add up to 10), then click “show keyword opportunities.”
- Read the results, keywords your competitors rank for that you don’t (hover over a position number to see the specific competitor page ranking for it).
- Filter to help make this readable using the upper filters for KD and volume, the lower filters to set ranking position ranges (e.g., competitor ranks top 10, you don’t rank).
- Click a specific intersection between two competitors to narrow the comparison further.
- Export the filtered output to your working sheet (as that export will become the input for clustering in Step 4).
When walking through this, your own judgment will be important. Especially when it comes to the second task where you select competitors. Competitor selection is a decision, and you need to pick “search” competitors. Those that you go up against for search visibility and not necessarily traditional competitors that your company talks about internally in sales teams, product teams, or other departments. The list of product/service competitors is often (if not typically) different from the list of search competitors.
Also, note that the raw output here is not a to-do list to get to. You’re going to likely find a big chunk being branded, off-intent, or not relevant. That’s another human call you need to make at this point.
Sticking with my third-party logistics firm example, the head term “3pl” sits at KD 47, 29,000/mo. That’s really tempting, but it’s difficult and broad. Meanwhile, “3pl fulfillment” is KD 11, 2,600/mo and “3pl companies” is KD 18, 5,000/mo, with much more defined B2B intent. I’m not saying you shouldn’t go big with goals, but when it comes to content, there are sweet spots to try to hit before you take on the top industry term.
See also: Why Search Volume Is Screening Out Your Best Content Opportunities
Step 3: Read The SERP, Not Just The Keyword (The AI Search Step)
Building on our work so far, it is important to note that volume and KD help determine whether you can rank. They don’t go as far as telling you what that ranking is worth. In an AI search SERP, the first organic spot can sit below an AI Overview, among other search features, including ads. So, the same position today can deliver a fraction of the clicks it used to. Ahrefs found the top-ranking page loses roughly 58% of its clicks when an AI Overview is present. That puts even more importance on knowing what the SERP actually is, including the prevalence of AI Overviews, before fully committing to a term.
The Ahrefs tasks for this step:
- In Keywords Explorer, open a keyword ideas report (e.g., “matching terms”) for your shortlist from Step 2.
- Apply the filter SERP features > Include > AI Overview (to isolate the terms that trigger an AI Overview).
- To see where you already show up, go to Site Explorer > Organic keywords and filter SERP features > Current > Include target in > AI Overview (this gets to terms where your pages are already cited, or not, inside AI Overviews).
- Click into the SERP overview for your priority terms to see the full feature set (AI Overview, People Also Ask, ads, local pack – and who’s being cited in the AI Overview).
- Record each keyword’s SERP features on the roadmap (SERP features column), so in Step 5, we can account for it.
Again, I’m spelling out some tactical steps, but where you need to take some time and make a judgment call is related to where you see AI Overviews showing up. This is not an automatic “skip” scenario. There’s a decision to be made per term on whether the AI Overview presence changes whether you want to target it or just how you write for it. Some logic to factor in: For informational, top-of-funnel terms where an AI Overview dominates, you’re not going to get as many clicks. Depending on what you prioritize, you might care about writing for it or instead focusing on the organic result that will appear below it in the SERP.
In the 3PL example, in our logistics roadmap, an AI Overview appeared on nearly every B2B term we looked at (“3pl companies,” “3pl fulfillment,” “third-party logistics companies”). That reality had an impact on our entire content plan as visibility on those terms means getting cited in the AI Overview and contextualizing the SERP expectations for performance and traffic overall.
See also: When AI Takes The Click, Click Worthiness Should Guide Your Strategy
Step 4: Cluster And Group
A keyword list isn’t a plan. You get value from this work and see it through when you match how AI systems actually read topical authority. That comes to life in building around clusters and topics, not one-off pages chasing single keywords. So, to help us get closer to prioritizing, we have to build topics.
The Ahrefs tasks for this step:
- In Keywords Explorer, open the Matching terms report for your seed terms and use the clustering view (Ahrefs can cluster by Parent Topic – keywords that share the same top-ranking page – or by shared terms).
- Use Parent Topic clustering to find where several keywords can be served by one page versus where they truly need separate pages.
- Bring the clusters into your working sheet and organize them (Cluster > Group > Keyword – the structure our roadmap uses).
- Working sheet columns: Cluster, Group, Keyword, KD, Volume, Intent, and SERP Features
- Map each cluster to a specific destination (a hub page, service page, or article) so each cluster has a respective location on the website.
In this work within this step, you have yet another judgment call to make. The tool itself clusters by ranking overlap. However, your focus is on clustering by business meaning. Sometimes Ahrefs will split terms you’d serve with one page or will group terms together that deserve their own pages. You need to intervene and make the call on whether a cluster maps to a real buyer intent/need and page you’d build, and not an arbitrary or statistical grouping from the tool.
Again, bringing this back to the 3PL company example, in our logistics roadmap, terms rolled up into clusters like Foundation > 3PL (the core service terms: “3pl companies,” “3pl services,” “3pl warehouse”) and a separate Compliance > Amazon FBA Preparation cluster (“amazon 3pl,” “amazon fba prep center”). Those are two different business buyer intents with different pages, even though the tool may see them as being adjacent. The FBA cluster is bottom of the funnel, specific need, and not a variant of the general “3pl” terms.
Step 5: Prioritize: Difficulty × Potential × Intent × AI Overview Reality
Now, we’re at the point where we get to decide what gets built and the order. We’re bringing together the four things we’ve gathered in our work in this process so far, including difficulty (can we win), potential (is it worth winning), intent (does it map to the objective), and the AI Overview reality (what’s a click actually worth). Tools are valuable, but no tool quantifies or scores this for you–this is the human layer.
The tasks for this step:
- In your sheet, line up each cluster’s keywords with volume, KD, intent stage, and the SERP features (AI Overview) note from Step 3.
- Score against the Step 1 objective first (proximity to a lead/revenue being more important than volume).
- Tier your roadmap with a Foundation set you build first (winnable + high intent), then everything else behind it.
- Flag terms where an AI Overview means you’re optimizing to be cited, rather than earning a click, so expectations (and the brief you’ll complete in Step 6) are correct.
As noted, this step is much more human decision-focused rather than research tool-heavy. Your reasoning needs to be solid for your plan to come together connected to your goals. A high-volume, low-KD term can still lose to a lower-volume term that sits closer to a lead conversion.
In our 3PL example, rather than leading with the tempting term “3pl” (which is broad, mostly informational intent, and AI Overview dominated), we prioritized the more winnable, higher-intent terms into Foundation – “3pl fulfillment” and “3pl companies.” We pushed the Amazon FBA prep cluster up the list because the searcher intent is closer to buying, even with fewer searches per month. Volume was the last tiebreaker here and not the first filter.
Step 6: Hand Off To The Brief And Write For AI Visibility
A roadmap proves valuable only when it turns into the content it is designed to create. And, with the current search landscape, that means it needs to be built with an understanding of AI Overviews and LLM answers with citations, not just to rank in blue links.
Tasks for this step:
- Turn each prioritized cluster into a brief (primary keyword, supporting cluster/terms, intent stage, SERP features note, target page, and angle).
- Write to be citable: Lead each section with a direct two- to three-sentence answer to the question, then support it. Use headings that mirror how people actually search while keeping answers concise enough to be lifted.
- Be specific in ways that AI systems reward (real expertise, concrete detail, named entities, data – things that are fast and cheap, but not good; AI-created pages can’t fake).
- Add the technical layer (schema/structured data, internal links from the cluster’s supporting pieces to the hub).
Naturally, in this last step, you’re making some final judgment calls as the tools have already helped us by doing their work. AI can accelerate the brief and writing of the first draft, but the real examples, expertise, point of view, and human editorial judgment are what create credibility and resonate with the human reader consuming your content. The “good” we’re going for is to protect ourselves from the “fast” and “cheap” of a competitor coming up with a quick strategy with AI-only content that can’t replicate our expertise.
Good, Fast & Cheap Starts With A Baseline
At this point, I think we’re closer to nailing the mythical trifecta of having good, fast, and cheap. You may need to take a step back with your roadmap, briefs, and work you’ve completed to get out of the weeds to see it, but I want to encourage you and congratulate you for how you’ve prioritized the “good” as you work to scale for fast and cheap again.
What changes when going through this approach (or borrowing any missing elements from it for your existing one) is our research for the AI SERP, not a traditional approach that only cares about the blue link SERP. We’re optimizing now for citation along with intent and position. With the focus of this process and article being on AI Overviews, remember that this roadmap can and will extend to LLM citation sources, and be mindful of how having a structured approach can help for LLM focuses, too.
As AI SERP features continue to dominate, the cost of skipping this discipline increases. Generating visibility without a business outcome or ROI component mapped to it is a newer way to waste time and actual dollars spent on SEO and AI optimization efforts, but it can be harder to catch because the impressions and rankings can still look good in analytics and reporting.
There are two things to do with this roadmap and work.
First, determine your baseline. Do it by capturing your current SERP feature and click data before AI features shift significantly again so you can track your progress from this point forward. The click details and math will look different now than a couple of years ago and are sure to look much different again in the future.
Second, don’t try to change your whole content operation at once. Test this method by picking one cluster, running it through these steps before you brief your next content piece, compare how it performs to something built the previous way, and use that learning to determine how you move forward with this (or any) framework.
More Resources:
- From Organic Search To AI Answers: How To Redesign SEO Content Workflows
- Why SEO Roadmaps Break In January (And How To Build Ones That Survive The Year)
- How To Create An SEO Roadmap That Actually Drives Results
Featured Image: Master1305/Shutterstock
