Your traffic is flat, and you don’t have a prosperity to throw at the problem. What should you test first? And how do you get your chief to endorse an experiment whenever you can’t commitment it volition work?
I put those questions to Jonathan Berthold, VP of Revenue at Ziff Davis, on The Search Engine Journal podcast. Jonathan is a longtime promotion strategist who plant in the two SEO and paid media.
We covered anywhere to discover bottom-of-funnel opportunities and how to keep a test small adequate that it can’t obtain downward a site. We additionally discussed how lengthy to delay before judging a result, alongside alongside a gap he sees in the two SEO and paid media: Marketers concentration on generating traffic and ignore what happens following a person lands on the site.
[00:45] What To Test When Traffic Is Flat And Budget Is Tight
[02:44] Getting Leadership Buy-In For Experiments
[06:35] Testing Small Before Making Sitewide Changes
[07:37] How Long To Run An SEO Experiment
[09:00] How To Evaluate Paid Media Tests
[10:51] Landing Page And Funnel Tests Marketers Overlook
[13:04] Finding Link Opportunities In AI Citation Data
[14:29] Why Search Intent Matters More Than Volume
Start With A Site Analysis, Then Target Bottom-Of-Funnel Content
When traffic is flat and there’s no budget, Jonathan starts alongside a complete location inspection to discover out why traffic stalled. He has inherited sites in that circumstance before, ranging from B2C SaaS companies to informational lead-generation sites. What he finds in the inspection determines the fix. Sometimes that method improving existing pages, and sometimes it method construction new pages to inhabit a topic gap.
Either way, his test would mark business intent. Top-of-funnel traffic is being “siphoned by LLMs, AI Overviews, AI Mode,” he said, and that traffic “is going distant and/or doesn’t really change or add to the base line.” So he would test one of two approaches: optimizing existing pages for business and transactional keywords, or creating new satisfied to inhabit topic gaps in the market.
Without a budget, Jonathan relies on AI to do the research. He uses it to study what category of satisfied ranks for bottom-of-funnel conditions and anywhere the opportunities are to measure traffic in areas competitors aren’t covering. He additionally uses it to fig out what added viewpoint would create the new pages “unique and inviting to hunt engines akin Google.”
I asked what test most marketers have to be operating but aren’t. Jonathan stated there is “a enormous market for bottom-of-funnel traffic that a lot of SEOs sleep on.”
It’s easier to build a scheme about keywords alongside additional than a thousand monthly searches, but many of those keywords are informational. Giants akin Reddit, LinkedIn and Wikipedia already position for them, and LLMs, AI Overviews and AI Mode have absorbed much of that traffic. He called bottom-of-funnel targeting “100% the easiest low lift opportunity” available.
Get Buy-In By Defining What You Will Learn
Jonathan sees endorsement as a communication problem. He never tells a chief that the squad volition test item and anticipation for a result. Every test is an chance to learn. He starts alongside a hypothesis, and the throw explains what the test volition reveal. It power display whether a maneuver or procedure volition assistance in the future, or whether a messaging or imaginative idea holds up.
He recommends framing the ask about the long-term payoff by explaining what the squad volition study and how that cognition volition create it improved at its craft. When the throw is positioned as “this test volition authorize us to study X, Y, and Z so that we can acquire this much into the future,” he said, it is “easier understood by a guidance squad that mostly plant inside P&L.”
Start Small So One Test Can’t Tank The Site
I raised a involvement concerning the quick fixes circulating now that newcomers are using to act SEO through AEO and GEO. Those strategies activity until the algorithm catches up. I’d fairly keep experiments contained than put an complete website at risk.
Jonathan agrees. When a squad runs an SEO test throughout a complete website and the test goes in the incorrect direction, “you container the entire organic,” he said. His regulation for the two SEO and paid media is to commencement alongside a extremely small section, afterward iterate and develop gradually as the results arrive in.
Paid media taught him the identical lesson. Tripling a prosperity in one day “ruins all the optimization,” he said, so budgets need to increase gradually all day during the scheme learns. In SEO, the equal error is rolling out a schema or content-writing test sitewide and tanking pages that already execute well.
Results additionally need to be replicated. A outcome should reiterate “so that it’s not fair a random result,” he said, until a form emerges that you can use to the remainder of your work.
Give SEO Time, And Set A Different Rule For Paid Media
When a test isn’t going the way you hoped, whenever should you stop? Years ago, the norm direction was to provision an SEO test six to nine months, particularly for a new site, but iterations move much faster now. I asked Jonathan how he decides whether to provision a struggling test additional period or end it.
For SEO, he gives an test concerning a month. He relies on Search Console and treats its data as directional. Google needs period to crawl the updated pages, procedure the changes and modify rankings before impressions and clicks move, so “definitely inside a week, you won’t see results.”
After 30 days, there is normally adequate data to see whether impressions are moving, and he would delay 60 days at most. He stated starting small helps here, too, since it keeps expectations in inspect and lets the remainder of the team’s activity continue during the test runs.
Paid media follows a distinct regulation since the feedback is immediate. He either reviews tests all two weeks or uses his preferred method, which is a expend threshold of two to three times the mark disbursal per acquisition. Once a test has spent that much, it has run its course, and the squad can either iterate on it or slay it.
The Test Most Marketers Skip Happens After The Click
Jonathan sees the identical gap in SEO and paid media. Marketers put nearly all their notice on generating traffic and ignore “what happens whenever a person really hits your website.”
Any leaf that already draws a lot of integrated traffic, or that a squad spends heavily to dispatch paid traffic to, is a fine applicant for split testing and chimney optimization. Landing leaf and chimney experiments should continually be “an on thing,” he said, since you can keep raising conversion rates at all phase of the funnel, “even by a few basis percent points.”
Use Citation Data To Find Outreach Opportunities
Jonathan is additionally operating his own test that connects bottom-of-funnel targeting alongside nexus building. He uses AI citation data “to unearth nexus construction opportunities.” He sees it as a way to place anywhere LLMs are heading and to build brand existence by earning the citations competitors have.
Many companies halt abbreviated of that step. A lot of them track AI visibility and observe which sites citation their competitors, he said, “but not a lot are taking the next step, which is reaching out to these publications.”
The payoff compounds. Pages that citation competitors for bottom-of-funnel keywords apt position for those keywords themselves and get cited frequently in LLMs. Earning a citation on those pages can dispatch traffic to your location and get your brand cited alongside your competitors, which he called “a additional holistic approach.”
Use Intent To Choose Keywords When Search Volume Is Low
Jonathan no longer puts much inventory in hunt volume. Search habits alter all day. If the environment and topic are right, a long-tail keyword can motionless be value targeting equal whenever Google, Moz or another phase reports it at one, 10 or zero searches a month.
Volume is helpful as a directional signal, particularly in paid media, anywhere it helps set expectations for the total addressable market. In integrated search, fragmentation changes the math. People hunt in additional ways than they used to, including inner LLMs, and LLMs run their own query fan-outs. That makes targeting “how your mark ICP searches” additional crucial than the figure in a keyword tool.
His example comes from his agency days. A client in the composites industry supplied fiberglass conduits for business projects. Its total addressable market was extremely small, and its keywords had extremely low hunt volume. But if 10 group searched a duration and two of them landed on the site, “very apt they’re in that market. That’s a half-a-million-dollar contract.”
That’s anywhere conversion charge comes in. Someone searching for the finest home loan rates volition apt end up on informational satisfied and likely won’t change on that site. A hunt akin “I need a plumber to fix my toilet” may display small hunt volume, but the individual typing it has an urgent need and elevated intent, so they are much additional apt to convert. As LLMs rotate into additional extensively used, he expects hunt quantity to be “not completely representative” of the phrases group use or the many fan-out queries a sole keyword can generate.
Test Where the Revenue Is
As AI answers assimilate informational searches, Jonathan says concentration on high-intent searches, equal whenever the quantity is low. He additionally recommends operating tests on the pages anywhere visitors decide whether to convert. A small test in those places can display a measurable endeavor outcome inside a duration or two.
That outcome additionally makes the next test easier to get approved. A test tied to conversions produces a figure that a guidance squad can use to create audio decisions.
More Resources:
- Test Like A Scientist: 6 Mistakes That Undermine Your SEO Experiments
- Why Search Volume Is Screening Out Your Best Content Opportunities
- How We Build An SEO Content Roadmap For The AI Search Era (Step-By-Step)
Featured Image: Anton Vierietin/Shutterstock
