Home  /  Blog  /  IV

Digital PR for SaaS companies

The instrument that earns links outreach cannot buy — and the one most SaaS companies attempt once, badly, and conclude does not work.

Digital PR  ·  13 min read

THE SEQUENCE01Find the number02Verify it03Publish methodology04Pitch writers05Passive accrual
The last stage produces most of the links and requires no further spend. The first stage decides whether any of it works.

Digital PR in a SaaS context means one thing specifically: producing evidence that journalists and analysts need, and cannot get elsewhere, then making it easy for them to use.

It is not press releases about funding rounds. It is not thought leadership. It is not a survey of two hundred people about how excited they are for AI. It is a number, from a defensible source, answering a question your industry has been arguing about without evidence.

Why it works when it works

Journalists writing about your category need statistics and have very few sources. Vendor blogs are unusable — self-serving and unmethodical. Analyst reports are paywalled. Government data is too coarse. So a writer covering, say, customer support benchmarks has perhaps two citable sources, and both are five years old.

Producing the third is a durable position. The links that follow cannot be bought at any price, because what is being exchanged is not money — it is evidence the writer needed.

The best measure of a digital PR asset is not how many links the launch produced. It is how many arrived in month nine, from people you never contacted.

Finding the story

Three sources, in order of preference.

1. Your own product telemetry

The strongest option, because nobody else has it. A scheduling tool knows when meetings actually happen versus when they are booked. A support platform knows real first-response times across thousands of teams. A payments product knows failure rates by method and geography.

The question to ask your data team: "What do we know that nobody outside this company knows?" Then: which of those things would a journalist covering our industry want a number for?

Anonymise and aggregate properly, and say so in the methodology. This is both an ethical requirement and a credibility one — a study that does not explain its privacy handling invites the wrong questions.

2. A panel survey

Where usable internal data does not exist. Three hundred or more qualified respondents is roughly the threshold at which trade press treats a finding as citable rather than anecdotal.

Qualification matters more than volume. Three hundred verified practitioners in your category beats three thousand from a general consumer panel, and a survey of your own customers is not research — it is a customer survey, and a careful journalist will treat it as one.

3. Public data, recombined

The cheapest route. Take datasets that exist but have never been joined, and join them. Job postings against funding data. Regulatory filings against headcount. This produces genuine findings at low cost and is under-used because it is unglamorous.

The test for whether a finding is citable

Write the headline a journalist would use. If it names your company, the finding is promotional.

Promotional: "Companies using Northwind resolve tickets 40% faster."
Citable: "Median first-response time across 12,000 support teams is 4 hours 12 minutes — up from 3h 40m two years ago."

The second is useful to a writer whether or not they have heard of you. That is the entire distinction, and it is the one most SaaS companies get wrong on their first attempt.

PITCH LIST, BY CONVERSIONWriters who covered it without data100Publications citing comparable studies64Industry newsletters38Analysts21
Group one is the highest-converting list you will ever build, and it is findable in an hour: search your topic, find claims with no source, note the byline.

Building the asset

Requirements, all of which sound minor and none of which are optional.

  • Ungated. Completely. A journalist on deadline will not complete a form.
  • Methodology stated plainly. Sample size, period, collection method, exclusions, and how the data was anonymised. Without it, a careful writer cannot cite you and a careless one will misquote you.
  • Headline numbers near the top, plainly stated. If extracting the statistic requires reading three paragraphs, they will use somebody else's.
  • Charts available as images. Publications embed images and credit sources. One of the most reliable citation mechanisms there is.
  • Raw data downloadable where you can. It signals confidence and it earns links from people who want to reanalyse it.
  • A suggested citation line. Tell people how to reference it. It works more often than it should.

The pitch

Short. Journalists receive dozens of pitches daily and read the first two lines of most.

Structure that works:

  • Subject line: the finding, as a fact. Not "New research from Northwind" — "Median support response times slipped 14% in two years".
  • First line: the number and the sample. One sentence.
  • Second line: why it matters or what changed.
  • Third line: a link to the full data and methodology.
  • Fourth line: an offer — a named expert available for comment, or the raw dataset.

That is the whole email. No preamble about admiring their recent article, no company boilerplate, no attachment.

Who to contact

PriorityWhoWhy
1Writers who covered this topic without dataThey needed this and could not find it
2Publications that cited comparable studiesDemonstrated citation habit
3Industry newslettersHigh relevance, fast turnaround, links in the archive
4Analysts covering your categorySlow, and their citations are durable
5Communities where the question recursRarely links, often visibility

Group one is the highest-converting list you will ever build, and it is findable in an hour: search your topic, find articles asserting something without a source, note the byline.

Timeline and expectations

PhaseDurationWhat happens
Data access & analysis2–3 weeksFind the finding; kill it if it is not interesting
Build & write2 weeksReport page, charts, methodology, press kit
Pitch window2–3 weeksMost direct coverage lands here
Passive accrualMonths 2–14Usually the majority of total links

Why first attempts fail

Four recurring reasons, in rough order of frequency.

The finding was promotional. The headline named the company. Nothing recovers from this.

The sample was too small or unqualified. Under three hundred, or drawn from a general panel. Careful writers check.

No methodology was published. The number may be perfectly sound; without the method it cannot be responsibly cited.

It was gated. Astonishingly common, and it removes almost the entire value.

Cadence

One story per quarter is the sustainable rhythm for a mid-market programme. More than that and quality falls; less and you lose the accumulating benefit of being a known source.

The compounding effect is the reason to persist. By the third study, journalists who used the first two open your emails. That relationship is the actual asset — the individual studies are how it gets built.

The AI-search argument

One more reason this instrument has become more valuable rather than less. Answer engines building responses about your category draw on sources they treat as reliable, and original research published with a methodology is close to the ideal shape of such a source.

Being the origin of a statistic that gets repeated across dozens of publications means being named when a model summarises the topic. That is not a link, it is not measurable in a backlink tool, and it is increasingly where the value sits.

The short version

Find a number nobody else has, from product data or a 300+ qualified panel. If the headline names your company, start over. Ungated, methodology published, charts as images. Pitch writers who covered the topic without data. Expect most links between months two and fourteen.

Ask what is in your data