Back to the Campfire Blog

Data Disagreements: Finding Alignment When the Data Will Never Be Perfect

Written by: Adrienne Kmetz Tags: data, analytics

Published: Jun 9, 2025

"No data is perfect" - Dana DiTomaso

We've heard it from the top analytics experts. We know that reliable attribution will always be just out of reach; because the buying journey is messy and emotional. 

If even the most data savvy know that we'll never have perfect attribution, how can we find solutions and have hard conversations that help us get what we need: Transparency, predictability, and better decision-making from insights?

Ugh

That feeling when the client opens the meeting with, “I’ve got a report here that shows traffic from organic is down. Why is that?” 

Great question!!1! Your data might show something entirely different; or you wanted to highlight something else entirely. This can come as a major curveball, especially when you’re trying to champion the business case for a stronger SEO program.

Executives and marketers are looking at different dashboards, definitions, and outcomes. And sometimes, they're looking at nothing at all ¯\_(ツ)_/¯ just their gut feel and previous experience. At Kickpoint Playbook, Dana has outlined an analytics maturity model that can show you where you are on the spectrum between chaos and total real-time integration. 

The perfect storm comes together when the data feels off, budgets are tight, and headlines about the economy make leaders nervous (plus the normal existential dread of working in a volatile industry like search). 

Add in the skepticism around Google's AI rollout, and revenue managers are sounding alarms as they extrapolate their trajectories for the next few years. 

When there’s a mismatch, it's often due to a mix of using different sources, and bias

• Data trust issues: Many CEOs don’t fully trust data – especially if they’ve been burned before. A Teradata study found over 40% of execs don’t believe in their company’s data – especially when it's AI-supported – due to accuracy or context gaps.
• Bias of all kinds: From confirmation bias, recency bias, frequency bias, to subconscious prejudice, we already have a narrative we're seeking to validate. Executives have strong instincts and long memories. If your data contradicts that, it’s going to be a harder sell.
• Lack of context: High granularity typically leads to more confusion. Stakeholders don’t need (or want) every data point. They want clarity on how this affects their top priorities – usually 3 to 5 core metrics like revenue, profit, leads, or efficiency.
• Emotional investment (aka sunk cost): No one likes being told their baby is ugly – especially if they pushed hard for a certain strategy. Tread carefully and prepare your talking points. 
• Trying to measure unmeasurable things: It's impossible to measure the impacts of hidden parts of the customer journey – the emotion of a conversation with a friend, the frustration of dealing with customer service, that moment when you became a "fan for life." 

 

“CEOs often don’t understand how their firms use data and analytics to drive decision-making.” (Forbes)

 

There's about to be a lot of pressure on data and PR teams

Rand's been beating the "attribution is dying" and "invest in your brand" drum for years. 

And now we're scrambling to define just how to measure Brand, in an effort to find what the ideal tactic is for "getting into" LLMs, as conversational search and recommendations take over ranked listings.

Digital PR teams typically report on mentions, backlinks, and DA on a regular basis. The challenge is being able to take those numbers and create a connection to revenue, which we used to (attempt) to do through increased rankings and traffic. 

Our partner Citation Labs reported in its 2025 Link Building Survey that:

"Only 40% of respondents said their link building reports effectively show value. That means the majority are executing campaigns without a reliable way to demonstrate return on investment – a critical gap when it comes to justifying budget and prioritization."

Prepare and align, to avoid data disagreements

Use shared language

• How does this help us get there? Instead of practicing your imaginary speech, practice some code switching. Connect to business-critical goals and translate your data into desired outcomes.
  • Like: at the current rate this project could result in 500+ incremental leads. 
• Repeatable taglines: Give a tagline they can remember, feels impressive, and they can repeat to their stakeholders. Instead of presenting a raw dataset, translate it into the achievement it represents: 
  • For example, organic is fully aligned with paid media will mean more to a CEO than we completed the organic page + paid lander optimization project

We asked Dana what she thinks in her SEO Community Slack group’s Campfire chat. She said: 

“I find that it really helps to try very hard to make sure that I’m using terms that they understand; and I walk them through how the data they’re looking at could be wrong.

I also try to focus less on right vs. wrong and more on ‘we’re reporting in different ways, but this data will give us more useful insights.” - Dana DiTomaso

Watch her follow up video: 

Anticipate the FAQ

• Validate the sources: Show your work with a short methodology and sources used. Every quarter, audit your APIs, date ranges, and filters so you know connections are working. 
• Use their sources of truth whenever possible so you’re working with shared data. 
• Acknowledge limitations: No dataset is perfect, but most are directionally significant and we can still find patterns (albeit with a grain of salt). Be the first to acknowledge gaps so they don’t use them as an excuse to dismiss all the findings. 
• Be prepared to answer any question, even if it’s with “Great question! I’ll follow up with you on that." 

Find common ground

Sometimes, CEOs are operating on instinct and insights that aren’t purely numerical, and that’s something worth acknowledging. 

Nico Brooks shared in the SEO Community, "I’ve also grown to appreciate how much wisdom and knowledge exists in organizations that isn’t reflected in the data.”

Look for that middle ground that allows both the data and their perspective to coexist:

• Ask open-ended questions: Giving them space to articulate their hesitation can reveal concerns you hadn’t considered.
  • What part of this feels off to you? 
• Run a small test: Rather than debating the same issues from week to week, propose a "cheap and cheerful" experiment:
  • Let’s test this for one quarter and see what happens. 
• Frame the risk: Sometimes, the best motivator isn’t potential success – it’s the cost of inaction.
  • If we do nothing, here’s what we risk. 

Changing the company’s relationship with data over time

Leadership dismissing reliable data could be a symptom of a bigger issue around how leadership interacts with information. Discussion, disagreement, appreciative inquiry, and retrospectives are a normal part of work. Conflict, obfuscation, or dismissing data entirely, is not. 

Becoming data-driven means acknowledging the trust issues and mitigating them early on so you can get aligned faster and skip to the good part – getting creative from what you find in the insights. 

• Get leadership involved early: McKinsey consultants are famous for socializing reports and ideas early, which takes away the element of surprise and therefore reduces the tension and emotion during the meeting. 
• Build dashboards: Leaders should bookmark key dashboards so they can see data in real time. Consult the departments you need to align with so they know you're working on reporting that help connect the dots between your two groups. 
• Data is neutral: Avoid weaponizing data. If something goes up, great! If something goes down, it's not a failure of someone's work or character – we simply ask, "What can we learn from this?"
• "I don't know, let's look into it" is a suitable response. 
• Show wins with whys: Close the loop by reporting on the successes, and the team's tactics behind them. Give credit to your team members and shout them out when you have a great example of using data in action. 

Lead with empathy, align to business priorities, and use shared language. And remember: The best data in the world means nothing if it isn’t understood, contextualized, and acted upon.