Most B2B accounts spread non-brand search budget evenly across keyword themes and call it a strategy. It is closer to a refusal to choose.
A keyword theme is a set of keywords describing the same buyer problem in roughly the same words. Funding tiers rank those themes by how well their leads match the customers you actually close and what those leads cost, then pay them in that order. Tier one takes the majority of non-brand budget, tier two gets a defined smaller share once tier one is covered, and a small fixed remainder buys information about terminology that has not proven itself. This is written for a B2B team running one search budget across four or more distinct non-brand themes.
What a funding tier actually decides
A tier decides how much money a keyword theme is allowed to receive this quarter and what has to happen before that number changes.
It does not decide how the account is built. You can run three tiers inside two campaigns, or one tier spread across six ad groups. Tiers are a rule about money and priority, not a blueprint for containers. Teams that confuse the two end up rebuilding campaign structure when all they needed was a funding order.
The reason to write tiers down is that non-brand budget decisions otherwise get made by whoever is looking at the account that week. A tier list survives the week. An opinion does not.
Ready for paid ads that pay off?
Book your free auditRank themes by lead fit and cost, never by search volume
The two numbers that should order your themes are the share of leads from that theme that sales would take a meeting with, and what one of those leads costs. Volume tells you how many people typed something. It says nothing about whether any of them can sign a contract.
Google's own documentation on this is blunter than most advertisers realize. Keyword Planner reports average monthly searches for "a keyword and its close variants," those historical stats are "only shown for exact matches" regardless of which match type you plan to use, and the help page states plainly that "your search volume statistics are rounded" (About Keyword Planner forecasts).
The Competition column misleads people worse. Google defines it as "the number of advertisers that showed on each keyword relative to all keywords across Google." That is a count of other bidders. A theme can be crowded because it is valuable or crowded because it is cheap and vague, and the column reads identically in both cases.
So the ranking input has to come from your own data. Pull ninety days of leads by theme, mark each one qualified or not using the same definition sales uses, and divide spend by the qualified count. That single number reorders most non-brand theme lists immediately, and it is the same discipline behind optimizing paid search for lead quality instead of form fills.
If you cannot produce that number yet, stop reading about tiers and go fix the qualified signal first. Offline conversion imports exist for exactly this.
The three tiers, and what belongs in each
→ Tier one: product-category themes. The words a buyer uses when they already know a product like yours exists and are shopping for one. These earn the majority of non-brand budget because they produce the highest share of qualified leads at a defensible cost. In most mid-market B2B accounts there are two or three of these, not ten.
→ Tier two: adjacent workflow and vertical themes. The problem the buyer has before they know your category is the answer, plus industry-specific or regulated variants of your core terms. Real intent, worse conversion rates, longer path to a meeting. These get a defined share, and only after tier one is fully covered.
→ Tier three: unproven terminology. Emerging category names, terms a small number of buyers have started using, wording a competitor is trying to popularize. This tier is not an investment. It is a subscription to finding out whether the language is real yet.
Notice what is missing. There is no tier for "themes we have always run." Inheritance is not a ranking.
Fund tier one to its ceiling before tier two gets a dollar
Covering tier one means buying essentially all the qualified demand available in those themes before diversifying, not giving tier one the biggest slice of a pie you divided in advance.
The sequencing argument is arithmetic, not philosophy. A theme that produces a qualified lead at a good cost will keep doing that for the next impression too, up to the point where you run out of people searching. Money spent on tier two before that point buys worse leads while better ones go unbought.
The common objection is concentration risk. It is a real risk, and the answer is not an even split. Ask instead what would have to change for tier one to stop working, then price that scenario. Usually the honest answer is a competitor entering the auction, which shows up as rising cost per qualified lead over weeks, not overnight.
Do not put two tiers in the same shared budget
Keep tier one and tier two in separate budgets, because a shared budget is designed to move money across the exact line a tier list is meant to hold. This is where most tier lists quietly die.
Shared budgets are genuinely useful. Google describes them as "allowing underutilized budgets to automatically reallocate to budget-capped campaigns" (About shared budgets), and that automatic movement is the entire point. It is also the exact behavior that erases a tier rule.
In B2B the campaign that absorbs the leftover money is almost always the one you wanted to cap. Broader adjacent themes carry more available impressions and cheaper clicks, so they soak up spare budget far faster than a narrow product-category theme ever can. Six weeks later, the tier list on the slide says seventy percent to tier one while the account spent forty.
The working pattern is simple. Share budget freely inside a tier, where the campaigns are supposed to be interchangeable. Put a hard wall between tiers, where they are not. Tier three in particular should hold its own small budget that nothing else can reach, because its whole job is to keep buying information nobody would voluntarily fund.
Write the promotion rule before you need it
A tier list with no promotion rule becomes a permanent ranking, which defeats the purpose. Decide in advance what evidence moves a theme up.
Two conditions, both required.
→ Enough data to be a signal. Google's learning-period documentation says it can take "up to around 50 conversion events or 3 conversion cycles" for a bid strategy to calibrate to a new objective (Duration of the learning period). A conversion cycle is defined as the time from click to conversion, and when you import conversions it also covers the time those take to report (Conversion cycle: Definition). In a B2B account with a long sales motion, three cycles is a season, not a sprint. If a tier-two theme has produced eleven leads in six weeks, you have an anecdote.
→ Qualified cost that beats the bottom of the tier above. Not "promising." Not "improving." A number that beats the worst theme currently sitting one tier up, measured the same way.
Promotion should also cost something. When a theme moves up, the weakest theme in that tier moves down. Tiers with unlimited seats are just a list.
The mirror rule matters as much. A tier-one theme whose qualified cost has climbed for two consecutive review periods gets demoted, and the money goes to the theme that earned it. Review quarterly. Monthly reviews in B2B are mostly reading noise.
When tiers are the wrong tool
Three cases where this framework should be skipped.
→ Two or three themes total. Tiering three things is an org chart for a rowboat. Fund them and read the results.
→ Single-digit monthly conversions across all of non-brand. At that volume there is no theme-level read to act on. Consolidate, feed one conversion action, and revisit when the numbers support a decision.
→ A deal so large that one closed customer pays for a year of spend. Concentrated high-value niches justify buying an entire small theme outright rather than ranking it against anything. Deal value beats efficiency math here.
Do this in the next two weeks
1.) Export ninety days of non-brand search terms and group them into themes by buyer problem, on paper, before touching the account.
2.) Join each theme to qualified leads using the definition sales uses, and calculate cost per qualified lead per theme.
3.) Sort that list, draw the line for tier one, and write the dollar figure each tier gets this quarter.
4.) Rebuild budgets so tiers cannot borrow from each other, sharing only inside a tier.
5.) Write the promotion and demotion rule in the same document, with the date of the next review on it.
6.) Only then revisit match types, which decide how a funded theme reaches queries and should never be the first decision you make.
The output is one page listing every non-brand theme, its tier, its quarterly number, and what would move it. If your team cannot produce that page today, the even split is not a strategy you chose. It is the one you defaulted into.
If you want a second read on how your non-brand themes are actually performing, that is a large part of what our B2B Google Ads work starts with.
