Performance-optimization: How to Combine Paid Strategies, Manual Labor and Automation systems

What to focus on when optimizing online advertising, and what tools are good for this: together and separately.

The number of clicks and CTRs are the most obvious and the most accurate metrics. However, they are far from the most reliable for evaluating the effectiveness of ads. It is much more important to know how many clicks had to be paid, until the customer purchased the goods, and how many times the advertising revenue exceeds the costs for it.

Performance-optimization helps to cope with this: improvement of indicators reflecting the real result of the campaign – conversion and profit. Independent analysis, intellectual site strategies and advertising automation systems – we figure out why each approach is needed and how to competently combine them.

Strategy of advertising platforms

Yandex.Direct supports a number of “smart” bidding strategies. Among them there are those that allow you to increase the effect of clicks. First of all, this is average conversion price : the system raises bids on ads and keywords, most often leading to a conversion. Now the user pays not just for the visitor of the site, but for his very concrete action – for example, purchase. In order for the strategy to work, you need to specify this action in the Yandex.Metrica counter, tied to the campaign.

Of course, it is not always possible to get the average price, exactly equal to the given price: there are many influences, including the actions of competitors. But understanding what exactly the budget is spent on is worth it – even with a small margin of error.

Another algorithm is the average application installation price . The principle is exactly the same, only with a specific target effect. The application must be connected to the tracking system so that Direct can receive information about its settings. Finally, the average return on investment (ROI) . The coefficient itself is calculated on the basis of any convenient indicator: for example, commodity prices or margins. Its value is passed through the same Metric counters, so it’s important to set them up correctly.

Google AdWords’s smart strategies are basically the same. The mechanism of the target cost-per-conversion analyzes campaign statistics and allows you to get the average conversion price as close as possible to the given. The necessary action of the user is tracked on the page by a special code. The algorithm of the target profitability of investments in advertising is similar . Each conversion is assigned a value-for example, the cost of an order is calculated from it and the ROI is calculated.

Google, however, adds one more option – a set of strategies target CPA, target ROI and CPA optimizer. This solution uses machine learning, so that the bet on each auction is the best. The optimization takes into account many factors: the user’s device, location, time of day, etc.

To apply automatic Yandex and Google strategies, it is necessary for the campaign to collect certain statistics on clicks and conversions. The requirements of the algorithms are different, but with the advertisement just launched the systems will not be able to work: they have nothing to base on.

The Independent Analysis

Sometimes it’s easier to improve the results of an advertising campaign than it seems. Let’s consider the basic useful methods, which do not require complex calculations and skills of the programmer. They can be divided into two groups: related to your site and ad text and related directly to the campaign settings.

Among the first, strangely enough, increase the speed of loading the page . Yes, the probability of conversion is significantly reduced if the site is loaded for longer than a couple of seconds. The bottleneck here is third-party libraries like jQuery: install the caching plugin available in most CMS, and they will not take so much time. The increase in the number of conversion elements on the page – such as widgets and subscriptions – will also play into the hands.

Another simple step for optimization – improving the ad text . A. Belousov in the well-known work “Optimization of contextual advertising” proves that this method is more effective than focus coverage. With independent rotation and A / B tests, determine what causes your customers more interest. The right page for each ad will transfer the lead to where it can buy the desired product, rather than the less interesting home page. The development of this idea is adaptive tab headings varying depending on the user’s request. A person will see in the title of the page exactly the words he was looking for; For this, there are systems that are compatible with CMS. In general, any increase in usability of the site is useful for performance-optimization.

The campaign settings should be more careful: a person can not compete with the car in terms of speed and accuracy of actions, besides, their result is not always easy to predict. The “Quick statistics” tab in Google AdWords shows ROI by days of the week – it allows to adjust the bet in real time on days with the highest score. If you sort the keywords by the value of the conversion and its price, you can remove keywords with a zero conversion and lower the bid for words with too high a price .

The use of “negative keywords” in Directory and AdWords not only makes it easier to list all the keywords, but also significantly increases campaign coverage: each word has extensions that are not so easy to guess. A good lifefac – disable AdSense ads for applications. They are often clicked randomly, and the conversion is very rare.

Automation systems of advertising

Services that automate online marketing offer a more structured and more powerful performance-optimization tool. With them, most often work as advertising agencies and large advertisers: most have a paid subscription.

First of all, automation systems make it easy to work with keywords. Many of them – for example, Adobe Media Optimizer, Marylin or Alytics – provide optimal assignment of rates for large volumes of words . There is an opportunity to get rid of the competition between their words and from words that rarely lead to conversion .

The main difference is in combining several sites for campaigning from one interface. Hence a number of opportunities that are more difficult to implement independently. For example, the optimization of advertising on many KPI inaccessible in the sites themselves: CPA, the number of conversions, the percentage of failures, etc. in addition to ROI, CTR, CPC.

Often, services support the work with scripts, which makes them available to people without the skills of a programmer. Creating automatic rules for high flexibility rates allows you to customize the display of ads for the specific business. It is possible to hold the ad on the given position without complicated calculations and manual adjustments. Audience targeting is supplemented with adaptive ads containing, for example, the time until the end of the promotion or a specific search query of the user.

The Art of Balance

How to choose the right strategy? You can give only general advice: if advertising is launched for a startup, while you are technically and mathematically shoddy – it’s worth starting with manual configuration. In the end, you understand your business better than the still poor statistics.

When the campaign is launched, but sensitive to changes, and the budget is limited, you can rely on the “smart” strategy of Yandex and Google. In this case, you will be confident in how much you spend, and that the average indicators will be close to what you want. Finally, work with several sites and a large budget is effectively coordinated by automation systems.

However, the real secret of digital marketing is the ability to correctly combine all three approaches. When optimizing the site and the texts of ads, nothing will replace the human eye of a marketer and a web designer. Improvements in this area are much safer than manually adjusting site settings. To achieve the desired conversion and profitability within one platform, their own strategies are quite suitable: machine learning dynamically reacts to changes in the behavior of users.

In cases where the optimization goals go beyond the standard KPI, and the channels are used a lot, the automation automation services come into play. As we see, the more multidimensional calculations are required, the greater the need for powerful systems producing them. It should never be forgotten that the ads themselves are ultimately written for people.


Be the first to comment

Leave a Reply

Your email address will not be published.