SEO traffic forecast to the site: calculation of potential attendance

Table of Contents:

Any website that offers products or services should be potentially interesting to the user. How do you know how much traffic you can attract to a site and where the limit of growth is without experiencing everything?

We tell you why you need to predict the potential number of visitors to your site and share our approaches to traffic forecasting.

Traffic forecast is needed to know how many visitors you can attract to your site from search engines if you use certain SEO tools.

Calculate the attendance of the resource is necessary both before the launch of the project and in the process of its development.

We do this every time for client projects before starting work on resource promotion and optimization.

Calculation of attendance at the start of the project

When you have an idea to launch some kind of internet project, you need to predict the potential traffic to:

• To find out if there is a demand in this niche and if the idea is even worth considering seriously;
• Analyze market capacity, niche potential in general and the project in particular;
• Find out how much traffic is seasonally dependent;
• Make a development plan and promotion strategy to get the highest quality traffic possible;
• Assume the dynamics of development under different SEO strategies and budgets.

Traffic forecasting in internet projects is like a business plan in ordinary life.

In addition to the winning points, you can immediately take into account and minimize the risks. If you do not do this at the initial stage, there is a chance of missing something important, because of which the project will fail or then you will have to redesign the finished site for a long time.

Traffic forecasting on existing projects

In order for the project to develop successfully, you need to focus not only on your resource, it is important to occasionally analyze the niche as a whole and changes in search engines: track changes in the subject, analyze demand, competitors and new trends in promotion.

From time to time on running projects, you need to make traffic miscalculations and forecasts to:

• See potential points of growth;
• Analyze current or previous promotional strategy, its effectiveness and results;
• Evaluate the quality of the traffic you receive and see possible omissions;
• Identify possible problems that reduce conversion or traffic numbers.

Traffic forecasting as the basis of promotion strategy

Turning to SEO specialists, many clients expect instant and impressive results from optimization. Forecasting potential traffic will help to see the real picture with approximate implementation dates, avoid inflated expectations and subsequent disappointment.

Before starting work with a new client, we assess the traffic potential of the site, after which we propose a work plan. Based on the forecast data, we can roughly say what results can be obtained over a certain period of time with a given budget.

How to assess the resource potential on client projects: approaches to traffic forecasting

All work is divided into 4 phases:

• Estimate demand
• Determine CTR.
• Make traffic predictions.
• Make recommendations and a development plan.

Demand estimation

• We will find out whether there are plans to change the range of products and services in the future, the geography of the current markets, whether there are plans to expand the geography, and so on.
• Collect a semantic kernel, taking into account the future plans and business goals of the resource. It is necessary to sort the queries so that the site receives as much targeted traffic as possible, not just thematic. To do this, you need to understand what information requests indicate that a person intends to buy something, but later, and which show that a person is looking for information out of idle curiosity.
• We collect historical data on the frequency of queries in the desired regions for 2 years from the History of queries in Google. Most often we parse data on a monthly basis, but if necessary, it is possible to collect query history on a weekly basis.
• We determine the seasonality of demand and assign a coefficient to each month. To calculate the seasonality coefficient, we determine the average value of demand for all requests for the year, then for each month we calculate the coefficient using the formula.
• We forecast the dynamics of demand for the next year or other period taking into account seasonality. To do this, we first build a linear demand forecast for the next 12 months, and then adjust the forecast for each month with the calculated seasonality coefficient.

What's important here

Semantic core. It is important to collect the most complete semantics, not to miss anything, but at the same time not to take anything unnecessary. If any queries are missed, the forecast data will be lower.

For example, you sell bicycles, tires, swimsuits or tourist tours. Demand for these products depends on the season, and if you build your semantic core during a downturn, SemRush may not have many relevant queries. If you add extra queries, the data will be overestimated.

Sometimes optimizers add too general queries to the semantic core. For example, "tours", "smartphones", "windows", etc. Such queries often have a huge frequency, and they distort the real picture of demand. When forecasting traffic, we exclude such queries from the semantic core.

Frequency data. This data is the basis for future calculation of potential traffic. Frequency data will help you build your future strategy and select the queries you need to focus on. We collect semantics with SemRush.

Seasonality. During the year, the demand for goods and services decreases or increases. This factor is important to take into account when calculating potential traffic.

Seasonality data can also be taken into account when building a promotion strategy. For example, for goods with pronounced seasonality, it is better to start promotion 4-6 months before the beginning of the period when there will be maximum demand for these goods.

Having analyzed the demand, we can advise you to start promotion in September-December in order to get results at the height of the season, by June-July. If you start work later, there is a risk of missing the season.

Whether demand is growing or not. Thanks to the demand forecast, we can see the dynamics of niche development. It is important to know whether there are growth or decline trends, it helps to predict traffic more accurately and understand whether it is worth moving in this direction at all.

Definition of CTR

CTR is the ratio of the number of conversions to the number of impressions in search results.

CTR shows what proportion of users who entered a query and saw the site in the output went to the site. Each search string has its own CTR value. The lower the position in the output, the lower the CTR.

This value can vary depending on the topic, type of query, attractiveness of the snippet, brand awareness, competitors and the number of contextual ads in the issue, etc.

• Tabular values

  These are generally accepted average CTR values without taking into account topics and other factors. This method is the most inaccurate and is used when there is no real data on a specific project and niche as a whole.

• Clickability data from Google Analytics

  Based on them, it will be possible to calculate the overall average CTR for the entire site. How close this calculation will be to the real picture depends on the amount of accumulated data. Usually, at the beginning of promotion, a site has a small query index, so it is easier to take the average values as a basis as more reliable.

• Determine the CTR of each query for different positions

  This is the most accurate option available. The more popular a site is in search, the more accurate the CTR. We can do this if there is a lot of data accumulated on the project.

What's important here

CTR and display data in search will help calculate the visibility of the site. It helps to predict traffic and also find possible problems with the attractiveness of snippets.

When calculating CTR it is very important to cut off all branded traffic and not take it into account, otherwise the data will be incorrect.

For example, monthly traffic on the site is 120,000 visits. The average CTR for queries is 5.5%. At the same time, 12,000 visits are branded traffic with an average CTR of 35%. If we recalculate everything without branded traffic, we will get an average CTR ≈ 2.22% and this figure will reflect the real picture.

It turns out that the average CTR of branded traffic will overestimate the overall CTR of the site in search. We would have overestimated the forecast by 2.5 times if we hadn't excluded branded traffic.

Traffic forecast

There are two ways to predict traffic:

• based on the predicted positions of each query;
• based on site visibility.

In our work, we use the second method, as predicting positions for each query is a labor-intensive and practically impossible task without a large amount of accumulated data.

Calculation of potential traffic based on the forecast of positions of each query

1. Determine the position of the site on all queries from the semantic kernel.
2. Work with each query separately:
   • Predict the complexity and expediency of promotion on this or that query;
   • Calculate the future position of the query for the next year or for another period;
   • Forecast traffic for each query using CTR.
3. Add together the prediction results for each query.
4. Take into account seasonality data based on historical data. We adjust the forecast depending on the demand in a particular period.

The result you get is the potential traffic you can count on.

Traffic forecasting based on site visibility

Site visibility is a characteristic that reflects the position of the site on the semantic core depending on their frequency. The higher the visibility, the better the site is promoted in its niche. Site visibility depends on the following factors:

• the number of queries for which the snippet is shown in the visibility area;
• the number of queries for which the site is in the top;
• frequency of queries for which the site is in the top.

How we predict traffic based on site visibility data

• Determine all available initial traffic excluding clicks on branded queries.
• Determine the current visibility of the site.
• We predict the increase in visibility based on our own experience on similar projects, site quality, budget and other parameters.
• Make traffic forecasts based on projected changes in site visibility with seasonality of demand.

What to do with this data?

A traffic forecast is needed before starting a partnership to:

• Find points of growth and see the potential for development;
• Realize that it makes sense to invest in SEO and the cost of promotion will pay off.

Based on the forecast, we suggest possible courses of action to help the forecast become a reality.

We can also tie the amount of payment to the forecast and offer various guarantees in case of failure to achieve the targets. For example, if we fail to deliver the agreed amount of targeted traffic from search, we will compensate for the lost traffic with contextual advertising.

We can also tie the amount of payment to the forecast and offer various guarantees in case of failure to achieve the targets. For example, if we fail to deliver the agreed amount of targeted traffic from search, we will compensate for the lost traffic with contextual advertising.

How accurate can the predictions be?

A prediction is always something approximate, it may turn out better, or it may turn out worse. There are many factors that affect the final traffic:

• Competitor activity. It is possible that new players will appear on the market or that those already present will also start to promote themselves actively. Then it will be more difficult to fight for targeted traffic.
• Changing search engine algorithms. Yahoo or Google may change something in the ranking principles and you will have to spend time to adjust to the new rules. The principle of rendition may also change.
• Site baselines: commercial and behavioral factors, usability, structure, and more.
• Other external factors. For example, in Yahoo, the number of ad units and the type of organic output depends on the query itself and the user's behavioral history.

Demand may change with the economic situation, or the content of a search query may change from commercial to informational or vice versa.