HubSpot State of Marketing 2025: What the Data Signals

The HubSpot State of Marketing 2025 report is one of the more useful annual benchmarks in the industry, not because it tells you what to do, but because it shows you where attention and budget are moving across a large sample of practitioners. Read it as a directional signal, not a prescription.

What stands out in the 2025 edition is not any single finding, but the pattern underneath them: AI adoption is accelerating faster than most teams are prepared for, short-form video continues to outperform most other formats on engagement, and the gap between teams using marketing automation well and those using it poorly is widening. The question worth asking is not what the report says, but what it means for how you operate.

Why Annual Reports Like This One Are Worth Reading Carefully

I have a complicated relationship with industry reports. Early in my career, I treated them like instructions. Later, after running agencies and sitting across the table from CMOs who had done the same, I started treating them more like weather forecasts: useful context, but not a substitute for looking out the window yourself.

The HubSpot State of Marketing report is one of the better ones because it draws on a genuinely large respondent base and tends to ask commercially grounded questions rather than chasing trends for the sake of it. But every report of this type reflects the biases of its sample. HubSpot’s audience skews toward SME and mid-market marketers, many of whom are already inside the HubSpot ecosystem. That is not a reason to dismiss the findings, but it is a reason to interpret them in context rather than applying them wholesale.

If you want to think more rigorously about how signals like this should inform strategy rather than replace it, the broader conversation around marketing automation systems is worth exploring. There is a lot of noise in this space, and cutting through it requires a clear framework for what you are actually trying to achieve. The marketing automation hub at The Marketing Juice covers that framework in depth, from tool selection through to operational deployment.

What the AI Findings Actually Mean for Practitioners

The report’s AI findings are the ones generating the most commentary, and with good reason. Adoption rates have moved sharply upward in a short period of time. But the more interesting question is not how many marketers are using AI, it is what they are using it for and whether those use cases are generating measurable commercial value.

The most common reported uses are content creation, email personalisation, and data analysis. These are sensible starting points. They are also the areas where the risk of producing generic, low-differentiation output is highest. I have seen this play out in agency settings more than once: a team adopts an AI content tool, output volume increases, and then six months later someone notices that organic traffic has flatlined because the content is technically competent but commercially indistinct.

HubSpot’s own thinking on how AI can improve customer experience is worth reading alongside the report findings. The argument there is that AI works best when it is improving the quality and relevance of interactions, not just the volume of outputs. That framing is more useful than the generic “AI saves time” narrative that dominates most coverage of this topic.

The practical implication is this: if your team is using AI to produce more content faster, you should also be measuring whether that content is performing better or worse than what you produced before. Volume is not a proxy for value. I have managed enough P&Ls to know that activity and output are not the same thing as commercial contribution.

The Short-Form Video Signal: Format Fatigue or Genuine Durability?

Short-form video has appeared near the top of the HubSpot report’s highest-ROI formats for several consecutive years now. At some point, that consistency stops being a trend and starts being a structural reality of how people consume content.

The nuance in the 2025 data is that the format advantage is narrowing. More brands are producing short-form video, which means the bar for standing out has risen. The teams reporting strong ROI from short-form are not just posting more clips, they are being more deliberate about the first three seconds, the platform-specific context, and the call to action. The format is mature enough now that “we made a Reel” is not a strategy.

I think about this the same way I think about paid search in its early days. When I was at lastminute.com, we ran a paid search campaign for a music festival and generated six figures of revenue within roughly a day from a relatively straightforward campaign. That was possible because the channel was underutilised and the competition was thin. Short-form video had a similar window, probably between 2020 and 2023, where doing it at all was enough to differentiate. That window has closed. Quality and consistency are now the variables that matter.

Marketing Automation: The Gap Between Adoption and Effective Use

The report’s findings on marketing automation are where I spend the most time, partly because it is an area I have watched closely across a lot of different organisations, and partly because the gap between what automation can do and what most teams actually get from it remains stubbornly wide.

The teams reporting the strongest results from automation share a few characteristics. They have clean, well-structured data. They have defined their customer segments before configuring their tools. And they have someone accountable for the automation layer who understands both the commercial objectives and the technical constraints. That sounds obvious, but in practice it is rare. Most automation implementations I have seen start with the tool and work backwards to the strategy, which is exactly the wrong order.

Personalisation at scale is cited in the report as a primary driver of ROI improvement among high-performing teams. That finding aligns with what I have observed across client work and agency operations. But personalisation is not just inserting a first name into an email subject line. It is sending the right message to the right segment at the right point in the purchase cycle, based on behavioural signals rather than demographic assumptions. Getting there requires both the right tooling and the right data architecture underneath it.

The report also touches on the challenge of fragmented tech stacks. This is a real operational problem. I have worked with organisations running eight or nine separate marketing tools with minimal integration between them, which means the data sitting in each tool tells a different story. When that happens, measurement becomes a political exercise rather than an analytical one. Teams end up defending the numbers from their own tool rather than looking at the combined picture. That is a structural problem, not a reporting problem, and it requires a structural fix.

Social Media ROI: Honest About Attribution, Finally

One of the more encouraging aspects of the 2025 report is that the conversation around social media ROI has matured. There is less of the breathless “social drives everything” framing that dominated industry coverage for most of the 2010s, and more acknowledgement that attribution across social channels is genuinely difficult, particularly for brands with longer purchase cycles.

The report shows that marketers are increasingly distinguishing between the role social plays in awareness and consideration versus its role in direct conversion. That is a more honest and more useful framing. Social media does a lot of work in the upper funnel that is real and commercially valuable, even if it is hard to trace directly to revenue. The mistake is demanding the same attribution precision from awareness activity that you would apply to a bottom-funnel paid search campaign.

I judged the Effie Awards for a period, and one of the things that experience reinforced for me is that the campaigns with the strongest long-term commercial results were almost always the ones that treated brand and performance as complementary rather than competing. The teams that got this right were not measuring every channel by the same metric. They had a clear view of what each channel was supposed to do, and they measured accordingly.

The thinking around content specialisation is relevant here too. Copyblogger’s argument for specialised content applies equally to social strategy: generic content targeting everyone tends to convert no one, while content built around a specific audience and a specific moment in their decision process tends to perform disproportionately well. The report’s data on social ROI by platform reflects this, with the highest performers consistently being those who have made deliberate choices about where to focus rather than trying to maintain a presence everywhere.

What the Data Says About Content Strategy in 2025

The content findings in the report are interesting partly for what they confirm and partly for what they quietly signal about where things are heading. Long-form content continues to perform well for SEO and lead generation, but the report notes that the quality bar has risen sharply. This is consistent with what Moz has documented around how Google’s Helpful Content updates have reshaped what ranks. Generic, template-driven content is losing ground to content that demonstrates genuine expertise and specificity.

The Content Marketing Institute has been making a similar argument for years. Their how-to guides and frameworks consistently emphasise that content strategy should start with audience understanding, not with a content calendar. The HubSpot data supports this: the teams reporting strong content ROI are those who have done the audience research first and built their content programme around genuine gaps in what their target customers can find elsewhere.

I built my first website by teaching myself to code because the budget to hire someone was not available. That experience taught me something that has stayed with me: the constraint of limited resources forces you to be specific about what actually matters. Teams with large content budgets often produce a lot of content that nobody reads, because the volume feels like progress. Teams operating with tighter constraints tend to be more deliberate, and often get better results per piece as a consequence.

The Measurement Challenge: Honest Approximation Over False Precision

The report’s findings on measurement and analytics deserve more attention than they typically get. The headline numbers on tool adoption and channel performance are the ones that get shared on LinkedIn, but the underlying data on how teams are measuring effectiveness tells a more complicated story.

A significant proportion of respondents report difficulty connecting marketing activity to revenue outcomes. This is not a new problem, but it is one that has become more acute as the number of touchpoints in a typical purchase experience has increased. The temptation is to invest in more sophisticated attribution tooling, and the report does show increased adoption of multi-touch attribution models. But attribution models are a perspective on reality, not reality itself. They reflect assumptions about how customers make decisions, and those assumptions are often wrong in ways that are hard to detect.

The more useful approach, in my experience, is to be honest about what you can and cannot measure, and to build your measurement framework around the decisions you actually need to make rather than the data you happen to have available. If you are running campaigns across a long purchase cycle, last-click attribution is not just imprecise, it is actively misleading. Acknowledging that and building a measurement approach that accounts for it is more commercially useful than producing dashboards that look rigorous but rest on flawed assumptions.

Tools like Hotjar can add a qualitative layer to quantitative data, helping teams understand not just what users are doing but why, which is particularly useful when conversion rates are underperforming and the numbers alone are not telling you enough to act on.

How to Use This Report Without Being Misled by It

The most useful thing I can say about the HubSpot State of Marketing 2025 report is that it is a starting point for thinking, not a conclusion. The findings reflect what a large sample of marketers are doing and reporting, which is valuable context. But what works for the median respondent in that sample may not be what works for your specific market, your specific customer base, or your specific commercial objectives.

Read the report with three questions in mind. First, which findings are consistent with what you are already observing in your own data? Those are the ones most worth acting on. Second, which findings contradict your experience or your instincts? Those are worth investigating further rather than dismissing. Third, which findings are based on self-reported data from marketers about their own performance? Those deserve the most scepticism, because people consistently overestimate the effectiveness of their own activity.

The broader discipline of marketing automation, and how it connects to measurement, personalisation, and commercial performance, is something worth investing time in understanding properly. If you are building or refining your approach to automation, the resources on marketing automation at The Marketing Juice are a good place to ground that thinking in practical, commercially oriented frameworks rather than vendor-driven narratives.

The report is a signal. What you do with it depends on your ability to read that signal in context, filter out the noise, and connect it to the specific commercial problem you are trying to solve. That is the skill that separates marketers who generate business outcomes from those who generate activity.

About the Author

Keith Lacy is a marketing strategist and former agency CEO with 20+ years of experience across agency leadership, performance marketing, and commercial strategy. He writes The Marketing Juice to cut through the noise and share what works.

Frequently Asked Questions