Author: Maeva Cifuentes
SEO sales were easier for my agency in 2020, when SaaS companies were lighting money on fire and I only had to speak to a content manager.
Today, I sell to the CFO and to a marketer who is terrified that any bad decision will cost their job. I can’t close a deal without showing them what I think the impact of our work together will be.
I can’t get an idea across the line without showing how it might tie into results. It’s better that way—I am more accountable.
Ultimately, this means creating an SEO forecast for prospective clients. The somewhat tricky part is that there are many ways to do this, and (as with any forecasting) the projections will be an educated estimate.
But the more information you have, the more accurately you can forecast, and the safer you can make your client feel.
In this article, I’ll show you how I forecast as an agency for my clients, both for sales conversations and for getting buy-in for ideas for existing clients.
Table of contents:
Define unique success metrics for each client
It’s easy to say that every client only has one goal: more revenue.
While ultimately, this is true, it’s also reductive.
Your clients’ goals will differ based on their industry, growth stage, target audience, specific market challenges, as well as their leaderships’ visions and pressures.
Understanding and addressing these specific factors helps you land (or renew) the potential client. Metrics that are unique to that client will highlight your success towards their goals. To that end:
Understand the client’s goals
Identify the POC’s goals
Address client expectations and data consistency
Understand the client business’s goals
Broader company goals can serve as starting points for how you want to think about the ‘story’ of forecasting. These include:
Growing revenue
Improving customer acquisition cost
Building an audience
Etc.
Remember, ROI models alone don’t make a business case. A business case is a story that connects an executive’s priority to a unique solution you provide. To that end, Nate Nasralla, co-founder of Fluint.io, says that there should be a specific, named initiative that what you’re selling is rolled into—otherwise, it will most likely get deprioritized.
For example, Nasralla mentioned that Jim Franklin (former CEO of Sendgrid) shared his phrase: “Make the Mail Move.” It’s a phrase executives repeat over and over. In this case, “whether you were working on user acquisition, deliverability, or a new product feature, it all tied back to email volume. More accounts, successfully sending more emails.”
If you can find out things like what the exec team says is an internal priority at the all-hands, what the top-down OKRs are, etc. you’ll better understand how to frame your forecasts with regard to their top concerns.
Identify the personal goals of your contact person
Your point of contact (on the client side) is bound to have their own personal goals. Most of the time this will align with company goals, but sometimes they differ and it’s your job to find out what they are.
They might care deeply about the quality and even the level of humor in the content you’re creating—but, this might not affect the top-level company goals at all.
If this is the case, you can emphasize forecasting metrics tied to organic traffic growth from high-quality content.
This not only reassures them, it also aligns your forecast with what they value most, building trust and increasing buy-in. To satisfy stakeholders at every level, build a forecast and pitch that bridges both the company goals and your contact’s personal goals.
Address client expectations and data consistency
Clients all want slightly differing things. However, as their SEO agency, you’ll have standard leading and lagging metrics that you need to track for your own understanding of whether the project is moving in the right direction. And you want to set clear expectations with the client.
There are a few things I always need to figure out at the start of a new client engagement. You’d be surprised to know that sometimes clients don’t even have the answers to these questions:
What are your department and company OKRs this quarter, and how should SEO contribute to that?
How accurately is your CRM pulling data from GA4 and vice versa? Will I see the same number of leads if I look in one or the other?
What is your current conversion rate from organic search, and what counts as a conversion?
Are there internal blockers (e.g., lack of resources, slow approval processes) that might affect how quickly we can execute the SEO strategy?
Most clients would love it if you could forecast demos, revenue, or pipeline. And while I want to say that SEO can contribute to these metrics, you’d be pulling these numbers out of thin air if the client:
a) Doesn’t have an accurate way to measure these metrics internally, and
b) Doesn’t have one to two years of historical data on how SEO contributed to the metrics
In the following sections, I outline methods to forecast traffic growth. After creating those forecasts, I make assumptions based on the client’s existing conversion rates and customer value to forecast revenue metrics.
When it comes to SEO for sales-led orgs, most revenue-related forecasts are informed assumptions—not quite a finger in the air, but it wouldn’t pass a peer review, either.
The safest type of SEO forecasting is a traffic forecast supported by a strategy showing how the traffic will be the ‘right’ kind of traffic.
You’ll want to agree with the client on whether traffic refers to clicks, users/new users, or sessions, as well as the time frame (e.g., 28 days, 30 days, exactly the days within the month).
Build a data-driven foundation for accurate forecasting
Most forecasts are built based on benchmarking (both industry-wide as well as historical performance). While clients value that SEO agencies have broad industry data, it's also crucial to analyze trends specific to their website to improve forecast accuracy.
For example, I recently made the mistake of forecasting a client’s growth based on where they were starting from (their DR, current traffic, backlinks, and investment) without considering past trends.
So while I based my forecasts on their resources and my experience from past clients, I did not account for the fact that they were on a downward trend—not an upward one. I’d have to close the leak before we can start growing again, adding several months to my actual forecast. You can see in the graph above there was another month or two of decline before we addressed it and the client started growing again.
The most logical and mathematically accurate way to create forecasts is to launch a multivariate regression analysis. However, this requires coding skills and a professional data analyst, which most SEO agencies (including mine) don’t have.
From what I’ve seen, though, more complex and mathematical forecasts aren’t necessarily more accurate than simpler ones (more on that below). Now, that doesn’t mean your forecast should be back-of-napkin math, but I’ve done 100% of my forecasting with Google Search Console or Ahrefs exports, benchmarking, and Google Sheets formulas.
I’ll get into the specifics of various methods shortly, but first, let’s look at how you should contextualize your potential client’s historical data for more accurate forecasting.
Reflect seasonal trends
Seasonality is probably the easiest part of historical forecasting to account for since you don’t need other data, like how much the client was publishing or building backlinks.
For example, in the above graph, you can see that every July to November there is an average 20% drop in traffic that then slightly recovers (although not fully—there is still a downward trend here that I’ll discuss below).
The screenshot above is organic traffic for one of my client’s competitors. As we looked at other competitors, we saw the same trend across all of them:
All of the competitors follow this trend. For everyone in the industry, traffic went up until July and then it dropped until November, where it started the cycle again.
Reflecting cycles like this in your forecasting shows the potential client that you have insight into their industry and can plan around seasonal peaks.
Account for historical downward trends
A downward trend can often be seasonal (as shown in the section above), but sometimes it’s just a sign that something has been going wrong.
It’s especially important to identify this at the sales stage of your client intake process because otherwise, you’ll oversell and underdeliver. I personally try to avoid doing too deep of an unpaid SEO audit, but some agencies do a lot of unpaid work as a part of the sales process.
Without making it overly complex, you can use basic Google Sheets skills* to create a growth model based on estimated impacts:
Historical baseline — Use the last six months of declining traffic to establish the baseline.
Initial dip or stagnation — Assume that traffic may continue to decline for the first few months of your SEO work (since it takes time to reverse trends).
Gradual increase — Project traffic growth based on benchmarks and conservative growth rates (e.g., 5–10% MoM increase after the initial decline). You can also use Ahrefs or similar tools to estimate traffic growth from recovering lost keywords or acquiring backlinks.
Adjust for budget — If the client has a larger budget, you may forecast a more aggressive recovery based on greater resource allocation to content production or link-building campaigns.
*I walk through how I use Google Sheets to forecast in a later section.
Look at previous clients or public case studies where similar traffic drops were reversed. You can use this data to project:
The average timeline for stabilization after a decline
The percentage traffic increase that is typical for the client’s level of resource investment
Then, in your forecast, you can set different scenarios:
Scenario | SEO Forecast |
Worst-case | If no corrective actions are taken or if initial SEO efforts don’t immediately make an impact, client traffic could continue to decline or remain stagnant. Consider the possibility that more severe issues (e.g., technical errors, penalties) might be contributing to the performance decline. |
Moderate | With consistent SEO work and investment, traffic stabilizes after six to nine months. Here, you assume that the decline was mostly due to a lack of SEO effort or falling behind competitors. |
Best-case | If the drop in traffic is due to fixable issues, like outdated content or easy wins with backlink building, SEO efforts could start showing positive signs within three to six months, with gradual improvements afterward. |
Use total addressable market (TAM) to inform your forecasts
TAM is one of my favorite ways to forecast because it provides my agency with what feels like the most data-driven approach to forecasting, especially when I’m data poor (as I often am during a sales conversation).
In this approach, I’ll examine the prospective client’s top and closest* competitors.
*In some cases, clients will consider a company a competitor even though they are way out of their league SEO-wise.
If I were forecasting TAM for Constant Contact, for example, I’d remove Mailchimp from the equation, since nearly one million of its monthly visitors come from branded searches, and they also rank highly for crazy high-volume and intent-vague keywords like [content marketing] (position 1, 171K search volume) and [roi] (position one, 64K search volume). Focusing on competitors like ActiveCampaign or Klaviyo provides a more accurate basis for forecasting.
After I remove the outliers, then I compare traffic per page:
Domain | Monthly traffic | Pages | Visits per page |
681,149 | 171,965 | 3.9 | |
368,952 | 9,518 | 38.7 | |
322,208 | 13,120 | 24.5 | |
243,739 | 8,152 | 29.8 | |
134,784 | 1,894 | 71 |
The average visits per page in this industry is 28. You can also see that Constant Contact is highly inefficient per page of content.
That brings us to a true traffic per page of 115, which is highly efficient.
Let’s say you have a solution to improve content efficiency. Based on the benchmarks above, you can project between 5–71 visits per page.
If you add 120 pages in a year (or 10 pages a month), the lower-end assumption for new traffic would be 600 visits (assuming 5 visits per page) and the higher-end assumption would be 8,520 visits (assuming 71 visits per page).
If we started from 681,149 (Constant Contact’s current traffic), that’s a year-end forecast of either 681,749 monthly traffic or 689,669 monthly traffic. Or, otherwise said, a 0.08% YoY increase or a 1.2% YoY increase in Constant Contacts total website traffic.
This tells us two things. We can either:
Count only traffic growth from the blog as a success metric rather than the website as a whole; or
Increase the investment to make a larger impact on growth
The higher end assumes you implement a plan to improve content efficiency (e.g., through better content, higher volume keywords, more internal linking or linkbuilding).
Keep in mind that this method makes SEO growth seem linear (and we know that it isn’t), but again, it can give you a ballpark estimate which I have found to be about 85% accurate.
Build your forecast in Google Sheets
I like the above method because it gives you an understanding of deliverables in, outcome out. But sometimes, you’re doing a lot more than just publishing new pages.
Your client could invest in reoptimizing existing pages, building backlinks, improving technical performance, and creating new content.
So, sometimes a per-page forecast won’t work well for you, but you aren’t quite yet ready to use machine learning or advanced methods for forecasting.
There are a few ready-made Sheets templates you can use for this:
My SEO forecasting template (shown above) is based on benchmark data based on size of investment and current monthly traffic. There aren’t all the multiple variables in it and it doesn’t account for how well you will do the actual work, but we do SEO really well and have found this forecast to be 85% accurate, which is pretty good for an SEO forecast.
Advanced methods for SEO forecasting
I'm no mathematician and I prefer simpler forecasting models, but there are ways to make your SEO forecasting even more data-driven. (In sales conversations, I don’t know if these complex, time-consuming and expensive models are better, though.)
These advanced methods are particularly useful when dealing with large datasets, long-term forecasting needs, or when a high degree of accuracy is critical for decision-making.
I'll give a high-level overview of some options, but you’ll probably need a programmer or data analyst to really get going with these techniques. Here are some advanced methods to consider:
Time series analysis for understanding past traffic trends and making predictions based on historical patterns
Machine learning regression models to predict traffic or ranking based on various SEO factors
Natural language processing (NLP) to analyze keyword trends and predict emerging topics
Time series analysis
Time series analysis tracks historical SEO data like organic traffic, conversions, or keyword rankings to forecast future performance. This method allows you to identify upward or downward trends over a specific period and forecast whether that pattern will continue. It can also help you predict the impact of specific SEO changes.
For example, let’s say you’ve been steadily publishing blog posts over the last year and see a clear pattern that, after each post, there’s a temporary traffic spike followed by a gradual decline. Using time series analysis, you can predict the potential impact of publishing a series of similar blog posts over the next six months. You would analyze the historical data on how past blog posts performed and use it to estimate how much traffic you might gain by maintaining the same publishing frequency.
You can learn more about this from SEO analyst Jess Peck and Patrick Stox on the Ahrefs blog. You’ll need to learn how to use ARIMA and Python for these methods.
Machine learning regression models
Regression models take forecasting to a deeper level by analyzing the relationship between multiple SEO factors and predicting future outcomes based on historical data.
Specifically, regression models help you understand how changes in one factor (like backlinks or keyword ranking) impact another (like traffic or conversions).
Here’s how it works: you input data like keyword rankings, domain authority, content length, and backlink profiles into a model. The model then analyzes how each of these factors contributed to past traffic growth.
Based on that analysis, it predicts how changes in those factors will impact traffic going forward.
For example, if you plan to acquire 100 new backlinks, the model can forecast how much traffic increase you might expect, based on the impact backlinks had on traffic in the past. This method is very useful when planning SEO strategies or setting expectations for client results.
While you’ll need to study machine learning to really learn how to do this, you can learn more about using linear regression for SEO from this Thatware article and a tutorial on simple linear regressions on this Github.
Natural language processing for keyword forecasting
NLP can analyze large datasets of search queries to predict which keywords or topics will become more popular in the future. By understanding trends in search behavior, NLP can help you forecast future keyword opportunities that may not even be on your radar yet.
For instance, NLP tools like Google’s Natural Language API can comb through massive amounts of search data, looking for emerging topics that are gaining traction. Imagine identifying a new trend in your industry six months before it goes mainstream—NLP can help you get ahead of the curve.
You can learn more about this in Ida Silfverskiöld’s insightful article about predicting trends with NLP.
Balance quantitative and qualitative insights for greater accuracy
Now once you have your forecasting graphs, whether fancy or simple, you need to put a story together for it to land with your potential clients.
I like to engage with clients and try to understand what factors might affect the pace of implementation.
For example, do they have the time and resources to dedicate to this? I’ve set up a forecast before where the point of contact was actually not responsible for SEO and had no time to dedicate to the project.
I wasn’t able to publish on time, the project failed and obviously did not align with the forecast.
There are also other factors. If the client does other work that supports SEO, those efforts may drive branded traffic and ultimately influence your projections. These activities include:
Social media campaigns
PR campaigns
Marketing collaborations with other brands
SDR outreach
Any other brand awareness motions
In parallel, if the client plans to do nothing and rely only on SEO, that will make the forecast more challenging as well.
Use SEO forecasting to drive sales and retention
Most SEO sales conversations today require some kind of answer to the question, “What am I going to get out of this investment, and how long is it going to take?”
And while many SEO agencies refuse to commit to some kind of number or forecast, it’s one of the few things that can give you a leading edge against competitor agencies.
Our clients—marketing leaders—know better than anyone that they’re constantly on the chopping block at work in today’s economy. They are more afraid to mess up than to miss out. When working with professionals who are excited about doing the right thing but terrified of doing the wrong thing, providing a forecast (even if you’re clear it’s just an estimate) is a big step in helping them feel they’re making the right decision.
This also goes beyond the sales conversation. Most clients that switched to my agency from another agency tell me that it’s because they didn’t know the ‘why’ behind what their previous agencies were doing.
They say it feels opportunistic, like the agency is plucking keywords out of air. When you explain what you’re trying to build and back it with what you think it’s going to bring them, the decisions feel clearer and you’re less likely to get client pushback.
Plus, it makes them look good in front of their bosses—like they are also making data-driven decisions because you’ve shown them that you are.
SEO forecasting is a cornerstone of long-term organic success
SEO forecasting is not just about making educated guesses or providing your clients with the numbers they want to hear. It’s about building trust, ensuring accountability, and setting realistic expectations based on solid data and trends.
By incorporating historical performance, understanding seasonal fluctuations, and using tools like regression models or NLP for keyword forecasting, you can provide clients with a roadmap that gives them confidence in their decision to work with your agency.
Remember, SEO is a long-term game, and while forecasts might not always be precise, they help guide the strategy, allowing for adjustments along the way.
The key is to balance transparency with expertise, ensuring that your clients understand both the potential and the challenges.
In the end, a well-communicated forecast strengthens your client relationships and sets the stage for sustainable SEO growth.
Maeva Cifuentes - CEO & Founder, Flying Cat Maeva is the founder and CEO of Flying Cat Marketing, an SEO and content agency driving growth with a holistic, revenue-based SEO approach for B2B SaaS companies in HR tech, martech, and salestech. Maeva is also a fractional CMO, marketing advisor, and certified confidence coach. Linkedin