SaaS Digital Marketing Agency: Diagnose the Bottleneck Before You Shortlist

SaaS Digital Marketing Agency: Diagnose the Bottleneck Before You Shortlist

A Series A software company signs a SaaS digital marketing agency in January. The agency is good. It publishes well-researched articles, fixes the technical debt on the site, and builds a keyword plan that ranks. By July organic traffic has roughly doubled and the pipeline has not moved.

Nobody was negligent. The company bought demand capture when its actual problem was that its category barely generated any search volume, because almost nobody knew the product class existed. Capturing a larger share of a small number is a real improvement and it was never going to solve that.

The reason this happens so often is that the phrase covers several genuinely different businesses, and the shortlist you build from a search results page will contain all of them without distinguishing between them.

What the label actually spans

Ask five agencies that describe themselves this way what they do, and you will get five answers that overlap only in vocabulary.

Capture existing demandPeople are already searching

• Search visibility and technical site work
• Paid search and comparison placement
• Conversion work on the pages that already get traffic
• Only works if the search volume exists

Create new demandNobody is searching yet

• Category and narrative work
• Content, social and community
• Events, podcasts and analyst relations
• Slow, hard to attribute, occasionally the only option

Reach named accountsYou know exactly who you want

• Outbound to a defined target list
• Account-based programmes
• Partner and channel motion
• Fast to test, capped by market size

Two more sit alongside those three and are frequently the actual constraint. Product marketing decides whether the proposition is comprehensible at all, and no amount of distribution rescues a message the buyer cannot parse. Lifecycle and retention work decides whether the revenue you win stays, which in a subscription business often matters more than acquisition and is almost never what an acquisition agency sells.

An agency that is excellent at one of these is usually mediocre at the others, and the homepage will not tell you which one it is excellent at. The engagement type they propose is a better signal than anything they write about themselves.

Diagnose first, in one afternoon

The useful question is not which agency is best. It is where growth is actually stuck, because the answer selects the discipline and the discipline selects the shortlist.

1. Step 1Traffic but no signups

   A conversion and positioning problem, not a distribution one

2. Step 2Signups but no revenue

   A qualification or product-fit problem; more volume makes it worse

3. Step 3No traffic and low search volume

   Demand creation or outbound; capture has nothing to capture

4. Step 4Good pipeline, poor retention

   An acquisition agency will not fix this and may deepen it

Running that diagnosis before taking calls changes the conversations completely. You stop being pitched and start testing whether the agency's proposal matches a problem you have already named. Agencies that are strong tend to prefer this, because it lets them decline a bad fit early rather than discover it in month four.

The single most common misdiagnosis in B2B software is the third one. A team looks at flat organic traffic and concludes it needs better content, when the underlying situation is that the total addressable search demand for what they sell is a few hundred queries a month. In that situation the only channels with access to the market are the ones that reach people who are not looking, and no quantity of publishing changes it.

Deal size decides more than anything else

The other variable that should govern the shortlist is average contract value, because it changes which motion can pay for itself.

At low contract values the economics only work with volume and low cost per acquisition, which pushes toward self-serve, product-led growth and efficient paid channels. Human-heavy motions cannot be afforded per customer.

At high contract values the arithmetic inverts. A small number of correctly chosen accounts is worth more than a large amount of traffic, and the cost of a person spending real time on one account is trivially justified. Search visibility still matters for credibility during evaluation and rarely originates the deal.

Between those poles is where most mistakes happen, because both motions look defensible and the company usually tries both at half strength. What actually works at each end, and what the crossover looks like in practice, is set out in B2B SaaS lead generation, and the channel-level version in outbound lead generation for B2B SaaS.

Specialist, generalist, or in-house

Once the bottleneck is named, three ways of resolving it are available and they trade off differently.

A specialist agency in the discipline you need is usually the fastest route to competence. They have seen the problem before, they have opinions, and the ramp is short. The cost is coordination: if you need two disciplines you are managing two vendors, and neither owns the outcome end to end.

A generalist agency solves the coordination problem and dilutes the competence. This is a reasonable trade when the company is early and the constraint is that nothing at all is happening, because breadth at moderate quality beats depth in one area that turns out to be the wrong area. It stops being a reasonable trade once you know where the bottleneck is.

Hiring in-house is the most expensive to start and the cheapest to sustain, and it is the right answer sooner than most companies think for anything that is genuinely core. The test is whether the capability is a one-off project or a permanent function. Rebuilding a website is a project. Knowing which accounts are worth pursuing this quarter is a function, and functions belong inside.

The hybrid that works most consistently is an in-house owner who is accountable for the number, with agencies supplying execution capacity in named disciplines underneath them. The arrangement that fails most consistently is an agency with no internal counterpart, because there is nobody on your side whose job is to notice when it is not working.

What working looks like at month three

Worth agreeing in advance, because the alternative is discovering in month six that you and the agency had different pictures.

For demand capture, month three should show movement in the leading indicators that plausibly cause the lagging ones, and a specific set of pages or terms that have moved. Revenue attribution at that point is not credible and asking for it invites bad measurement.

For demand creation, month three shows almost nothing, and an agency that promises otherwise is either misrepresenting the discipline or planning to substitute a faster channel and call it the same thing. The honest deliverable is a body of work in market and a defensible narrative, judged on whether it is any good rather than on what it produced.

For outbound to named accounts, month three should show real conversations with real companies, because this is the one discipline with a genuinely short feedback loop. If a targeted outbound programme has produced nothing in three months, something is wrong with the list, the proposition or the delivery, and the diagnosis is available immediately rather than after a year.

That asymmetry in feedback speed is itself a reason to run outbound early even when it is not the long-term answer. It tells you within weeks whether your proposition lands with a segment, which is information the slower disciplines cannot give you and will happily spend a year not giving you.

What to check before signing

Five things separate an engagement that works from one that produces monthly decks.

Who does the work. Ask for the names of the people who will be on your account and how much of their week you get. Agency quality varies more within a firm than between firms, and the team on the pitch is often not the team on the account.

What the first ninety days produce. A plan is not a deliverable. Ask what will exist at the end of month three that would still be useful if you parted ways, and treat a vague answer as information.

How performance is defined. Traffic, impressions and engagement are inputs. If the agreement measures inputs, that is what you will receive, and the reporting will be accurate and useless.

What happens to the assets. Domains, ad accounts, content, sending infrastructure and data should be owned by you and portable. This is easy to agree at the start and expensive to argue about at the end.

Where the incentives point. Percentage-of-spend arrangements reward larger budgets. Retainers reward stable effort rather than results. Performance arrangements reward whatever the defined outcome is, which is why the definition is the whole negotiation. The shapes and what each one distorts are covered in marketing agency pricing models, and the specific case in performance-based marketing agency.

Shortlist test for a SaaS marketing agency

• ✓Yes: They asked about ACV, sales cycle and churn before proposing anything
• ✓Yes: They named a discipline rather than offering all of them
• ✓Yes: They can describe a client they were wrong about, and what they changed
• ✓Yes: The proposal maps to a bottleneck you had already identified
• ✓Yes: Assets and accounts are yours and portable from day one
• ✕No: The pitch covers SEO, paid, content, social, email and outbound equally
• ✕No: Success is defined in traffic, impressions or engagement

The third line is worth dwelling on. An agency that has never been wrong has either not been operating long or is not telling you about it. The useful version of that answer is specific: here is the segment we thought would respond, here is what happened, here is what we changed and how quickly.

On the outbound slice specifically

If the diagnosis points at reaching named accounts, one practical note about how that work should be run, because it is where the most damage gets done on a software company's behalf.

Our own position is one message per campaign, built on one premise, sent once. If a different premise is worth putting to the same account later, that is a separate campaign with its own reason to exist. For a SaaS company the argument is partly about the asset and partly about the market. The sending domains an agency uses on your behalf carry your name, and repairing a burned reputation takes weeks during which nothing sends. More importantly, the total number of companies that could plausibly buy a specialised B2B product is often in the low thousands, and a badly targeted programme can spend a meaningful fraction of that population in a single quarter.

That constraint is worth writing into the agreement rather than assuming, alongside the assets clause. What the work costs at various shapes and scales is broken down in what a B2B lead generation agency actually costs.

If you would rather see a target list and a first message built against your own market before choosing an agency at all, see what a first campaign looks like.