Exploring the Link Between Ahrefs DR and Majestic TF/CF: An Analysis of 100,000 Domains

In this research, we share our findings on the correlation between the metrics from two popular link profile evaluation tools, Ahrefs and Majestic. Our goal is to help SEO specialists leverage these metrics together for more effective domain evaluations and data-driven decision-making.

Analyzing a website’s link profile is crucial for success in search engine optimization. SEO professionals commonly use Ahrefs and Majestic to assess domain authority.

Initially, our team relied solely on Ahrefs, but we noticed a concerning issue: the data pulled from the API often didn’t align with what was available through the web interface. This raised questions about the reliability of Ahrefs’ metrics, prompting us to seek out an additional analytics tool. After researching the options, we chose to incorporate Majestic into our toolkit.

While working with both services, we discovered that their databases didn’t always match; links accounted for by one service were sometimes missing from the other. Moreover, it was unclear how to effectively compare Ahrefs’ Domain Rating (DR) with Majestic’s Trust Flow (TF) and Citation Flow (CF) for the same domain.

To address this, we conducted a study involving 100,000 domains to identify any correlations between DR and the metrics from Majestic. Our objective was to propose a method for evaluating DR based on CF and TF.

What is Ahrefs DR?

Domain Rating (DR) is a metric from Ahrefs that evaluates the strength of a domain’s link profile on a scale of 0 to 100. The higher the DR, the more authoritative a domain is considered in terms of its link profile. DR is calculated based on:

• The number of unique domains linking to the site.
• The authority of those linking domains.

DR is widely used in SEO to assess a website’s potential ability to rank in search engines. However, DR can be artificially influenced, for example, by purchasing many low-quality links. This raises concerns that Ahrefs’ assessment of domain authority may be outdated and susceptible to manipulation.

What are Majestic TF and CF?

Majestic offers two key metrics for evaluating a link profile:

Citation Flow (CF) — This metric indicates the number of links pointing to a site, regardless of their quality. A high CF suggests a large volume of link activity, but it doesn’t provide any insight into whether those links are authoritative or spammy.

Trust Flow (TF) — This measures link quality and is based on the level of “trust” assigned to links from authoritative and verified sites. As Majestic explains, there are “seed sites,” which are carefully selected trusted websites. The quality of links is assessed based on the number of links from these seed sites and their distance to the target site.

A high TF indicates that the site receives links from reputable sources and links to quality sites.

TF is generally considered more resistant to manipulation compared to CF and DR.

The ideal balance between Trust Flow and Citation Flow occurs when a domain has a substantial number of links, all of high quality.

How we encountered the problem of metrics inconsistency

As previously mentioned, we relied on Ahrefs’ Domain Rating (DR) for a long time to evaluate domain quality. However, several issues became evident over time:

• The data obtained through the Ahrefs API often differs significantly from what is available in the web interface.
• Ahrefs DR can be artificially inflated.
• The relevance of the Ahrefs link database can fluctuate; links may “fall off,” but these changes might not be reflected in the tool immediately.

In search of a more reliable and consistent metric, we turned to Majestic’s Trust Flow (TF), which has shown to be less susceptible to manipulation. This conclusion is supported by industry colleagues and reinforced by our experience using both tools.

Our findings suggest that a high TF score better reflects the overall quality of a resource and its link profile. For instance, if we compare two sets of resources with high Ahrefs DR and high TF, we found that the group with high TF tends to contain more genuinely high-quality resources.

While working with these metrics, we observed discrepancies in the data from Ahrefs and Majestic for the same domain. A site might display a high DR but have a low TF, and vice versa. These inconsistencies complicate domain assessment, create uncertainty about the quality of the link profile, and hinder effective decision-making in SEO strategy.

For example, we previously relied on Ahrefs’ DR metric to filter potential dropped domains for building PBN networks (we excluded domains with a DR of less than 10). However, for Majestic metrics, we found it challenging to determine what TF and CF values would correspond to an Ahrefs DR of 10.

Where We Looked for Answers and Why We Decided to Conduct Our Research

We decided to investigate what the market, official resources, and the SEO community had to say on this topic. However, we found very little information.

Most existing research focuses on analyzing individual metrics from various services, such as Ahrefs, Semrush, MOZ, or Majestic. There are only a handful of studies that compare and establish correlations between these indicators. Official resources, like the Ahrefs and Majestic blogs, provide detailed explanations of their metrics. While these materials are useful for understanding the principles behind each indicator, they do not offer insights into how these metrics relate to one another.

Additionally, articles on authoritative SEO resources often discuss the functionality of Ahrefs and Majestic tools, highlighting their advantages and disadvantages in a general context. However, we did not find a comprehensive comparative analysis of their metrics based on statistical data.

In professional communities and forums, such as Reddit and specialized SEO forums, discussions about the use of DR, TF, and CF metrics often arise, with users sharing their observations regarding discrepancies between indicators. However, these discussions tend to be subjective and lack large-scale research or empirical data, making it difficult to draw any firm conclusions.

In summary, while there are numerous resources on individual metrics and tools, few address their direct comparison and interrelationships. To fill this gap, we conducted our own research.

Our Goals:

• Understand if there is a relationship between DR, TF, and CF, and, if so, quantify it.
• Determine if DR can be estimated based on Majestic metrics.
• Provide SEO specialists with a tool for more accurate domain evaluation.

How the study was conducted

We collected data on 132,363 domains, randomly selected from our proprietary database, which contains tens of millions of domains. This database is used to aggregate information for internal SEO and PBN network research. To ensure the comparability of the data, we recorded the Domain Rating (DR), Citation Flow (CF), and Trust Flow (TF) values for each domain on the same day. All domains were included in this analysis, even those with one or more metrics equal to zero. For our analysis, we utilized the pandas, scipy, and statsmodels libraries.

During the analysis, we immediately observed the distribution of these metrics. We found a significant number of domains with a TF value of zero, while DR and especially CF often had non-zero values. This discrepancy can be attributed to the fact that TF is more sensitive to the quality of links, whereas CF reflects their quantity.

To assess the degree of correlation between the metrics, we calculated both the Pearson and Spearman correlation coefficients. Both methods indicated a strong positive correlation among all three indicators. Notably, the relationship between TF and DR was the strongest. This suggests that TF serves as a good predictor for DR, indicating a linear relationship between the two metrics.

Scatter plot of Majestic TrustFlow and CitationFlow with color indicating Ahrefs DR values.

We decided to test this assumption by performing regression analysis. By using CF and TF as independent variables and DR as the dependent variable, we built a linear model. The results were promising: the model explained over 70% of the variance in DR, indicating a high level of efficiency.

Based on our regression analysis, we derived the following formula to estimate DR:

DR = 0.86 × TF + 0.6 × CF + 0.7

This formula enables us to estimate the DR value based on known CF and TF. For example, if a domain has a CF of 10 and a TF of 5, the expected DR would be calculated as follows:

DR = 0.86 × 5 + 0.6 × 10 + 0.7 = 4.3 + 6 + 0.7 = 11

It’s important to note that the actual DR may not always match the calculated value, particularly if the domain has few links. As mentioned earlier, the services have different link databases that may not overlap. In other words, the same domain in the Ahrefs database may have entirely different links than those in the Majestic database. Therefore, this formula should be used as a guideline to estimate the approximate equivalent of Ahrefs’ DR based on the link profile available in the Majestic database.

Recognizing that a linear model may not account for all data characteristics, we also explored whether machine learning methods could enhance prediction accuracy. We tested several models, including Ridge and Lasso regressions, decision trees, random forests, and gradient boosting. These approaches allow us to consider possible nonlinearities and interactions between variables.

Structure of a regression tree built using the features TF and CF.

The results indicated that the nonlinear models only slightly improved the quality of the predictions, increasing the explained variance of DR to around 73%. This confirmed our initial assumption that the relationship between the metrics is predominantly linear, suggesting that employing more complex models does not offer substantial benefits.

Importantly, Trust Flow (TF) consistently emerged as the most significant factor in predicting DR across all models. This underscores that link quality, as indicated by TF, plays a crucial role in Ahrefs’ domain evaluation.

Our study demonstrates that, despite the differing methodologies and databases of Ahrefs and Majestic, their metrics are closely related. This insight can be valuable for SEO and digital marketing professionals, enabling them to better understand how to effectively utilize these tools together and accurately interpret their metrics.

Conclusions

1. A High Correlation Between Metrics

We found a significant correlation among all three metrics: DR, CF, and TF. The Pearson correlation coefficients, reaching up to 0.8 between DR and TF, indicate a strong linear relationship, particularly between these two metrics. Although the Spearman correlation also demonstrated a strong relationship, it was somewhat less pronounced, suggesting some partial non-linearity in the data.

2. TF as the Main Predictor

Trust Flow emerged as a more significant predictor for DR compared to Citation Flow. This aligns with the notion that TF reflects the quality of links, while CF emphasizes their quantity. During regression analysis, we derived a DR prediction formula in which the weight of TF was notably higher than that of CF, confirming the greater importance of TF in assessing domain authority.

3. DR Prediction Model

The constructed linear regression model indicated that CF and TF can predict DR with an accuracy of about 70% of its variance, which is a positive outcome. However, approximately 30% of the variance remains unexplained, likely due to other factors not accounted for in the model. Nonetheless, the resulting formula is a valuable tool for preliminary DR assessment based on Majestic data.

4. Prediction Limitations

An important consideration is the difference between the Ahrefs and Majestic databases. These services may collect data on different links, meaning that a domain with high TF and CF in Majestic may not necessarily have a similarly high DR in Ahrefs. This limitation should be taken into account when using the evaluation model, particularly for domains with low link mass.

5. Nonlinear Models Did Not Provide Clear Benefits

We tested several machine learning methods, including polynomial regression and decision trees, but found no significant improvement in prediction quality compared to the linear model. This indicates that the relationships among DR, CF, and TF are predominantly linear, despite minor deviations from this assumption.

6. Practical Application

The model can be useful for SEO specialists looking to quickly evaluate a domain without access to DR data from Ahrefs. The formula based on CF and TF can serve as a reference point for domain evaluation within promotion strategies and while working with PBNs. However, it’s crucial to remember that the accuracy of these predictions is not perfect and may vary depending on the database used by the services.

Applying the Research Results

Building a Link Profile

For instance, if you’re planning a link-building campaign for your client and using Ahrefs DR, you might find two potential sites for guest posts: Site A with a DR of 80 and Site B with a DR of 75. Based on the traditional metric, Site A appears to be the more attractive option. However, when you apply the new metric, you discover that Site B has a better and more relevant link profile, less spam, and a more engaged audience.

Identifying Toxic Links

If you notice a drop in your site’s search engine rankings and decide to conduct an audit, relying on Ahrefs DR may not reveal any obvious issues, as most backlinks come from domains with high DR. However, using the new metric calculated from Majestic data reveals that many of these domains have high levels of spam and low-quality content. This insight enables you to identify and reject toxic links in a timely manner, ultimately improving your site’s search engine ranking position.

Analyzing Competitors More Accurately

If you’re trying to understand why a competitor is attracting more organic traffic than your client’s online store, even with similar DR scores on Ahrefs, applying the new metric can provide clearer insights. You may find that the competitor has higher-quality backlinks from thematically relevant sites and better internal optimization. Armed with this data, you can develop a more effective SEO strategy to enhance your client’s performance.

Conclusion

Our study demonstrated that Ahrefs’ DR and Majestic’s TF/CF metrics are closely related, and it is indeed possible to estimate DR from Majestic data with a high degree of accuracy. Trust Flow emerged as the primary metric influencing DR, which aligns with its emphasis on link quality. However, the study also highlighted some limitations, including potential inaccuracies for extreme values and discrepancies in data between Ahrefs and Majestic.

The findings from our research have already been implemented in practice. By using the estimated DR score from Majestic to filter domains, we increased the percentage of quality domains from 15–20% to 25–30%. This improvement not only saved our staff time but also increased the number of sites lifted by PUB.

We hope that our study will encourage more informed usage of metrics within the SEO community and pave the way for further research in this area.

Written by Ivan Aleshin — Data Engineer at Traffic-Hunters.com