Yuma 4×4

Media and Communications

✅ Performance-маркетинг. Юнит-экономика

✅ Performance-маркетинг. Юнит-экономика

Hello friends! My name is Mikhail Snitko and the theme of
this video is performance marketing. Each ruble invested in advertising should
be profitable, but what if the typical marketing approach does not produce the desired results? The performance approach comes to the rescue. What is performance marketing? This is an approach based on data-based decisions. What is the problem? Unfortunately, most entrepreneurs do not support
their decisions with data from the business, as a result, decisions are made from the bulldozer. How is performance marketing good? The fact that it can be calculated, and not
just calculated, but expressed in monetary terms. Acting in conjunction with business goals,
performance marketing offers to measure the effectiveness of marketing in profit. Where do our customers come from? Customers in online marketing appear from
site visitors through conversion. What is the limitation? The fact is that all users who get to our
site are divided into two large classes. For those who have not yet made a purchase
from us, these users buy from the carrier neither a product nor a solution to their
problem, they are trying to buy from us the belief that our product will solve their problem. In fact, people buy faith in our product. Those people who have already bought from
us, they buy exactly the solution to their problem, which they are going to solve with
the help of our product. It is imperative to separate these user cohorts,
as their decision-making scenarios are completely different. In fact, if we know how much we spend on each
user who gets into our funnel and calculate how much he brings money to the cashier, then
we can calculate the amount of money brought by the business from the flow of users. But, in order to make decisions consciously,
it is important to consider that the cost of attracting a client is very dependent on
other values. We have two key metrics – these are the costs
of attracting a client and the income received from him. However, this pair is not very convenient,
because the cost of attracting a client depends on many characteristics. For example, how good our product is, how
well we retain our customers, what costs we have to retain them, etc. The simplest characteristic is conversion,
i.e. we must first attract the required number of users from the advertising channel, then
with the help of our product, convert these users into customers, but from the point of
view of decision-making it is not very convenient. Because, when we are faced with the task of
reducing the cost of attracting a client, we will think, and what is more important
for us, increase conversion, reduce marketing costs or increase, for example, the frequency
of user returns. In fact, one value depends on different metrics. That’s exactly for this, we are making the
transition from the client coordinate system to the user coordinate system. Where we compare the costs of attracting the
user and the income received from the attracted user. Therefore, we separate these two processes. The process of attracting users and the process
of converting users. Why is this transition necessary? In order to separate marketing and product
decisions among themselves, so that each decision is made separately. Further, based on the analysis of metrics,
we look for ways to improve site conversion and how to increase the return on current
advertising campaigns using conversion optimization. What is conversion optimization? This is a set of measures applied to advertising
campaigns in order to obtain, from each advertising channel, the maximum number of leads for minimal
money. With a typical approach to analytics, optimization
of advertising campaigns ends, based on the findings of the first conversion. The second conversion, namely, the ratio of
the number of applications to the number of sales, is often not analyzed at all. In fact, conversion is the most unlucky metric. Why? There is no metric over which analytics systems
and common sense would not mock more than conversion. What is the specificity of the conversion? Conversion gives a double effect. Which one? It increases the number of customers and at
the same time reduces their value. Due to the high conversion, we can redeem
more expensive traffic, thereby squeezing the auction. But historically, it was easier to determine
the cost of a lead than to calculate the conversion by channels. What is the problem? The fact is that if we have several advertising
channels involved, then looking at the cost of lead, it is impossible to understand which
advertising channel is more effective. The fact is that the cost of attracting users,
the conversion to the application, and the conversion from lead to client affect the
customer’s value. These are all different metrics and each of
them has its own class of problems. Therefore, when we say that we increase conversion,
in fact, we reduce the number of those who did not buy. But often the numbers in analytics are tightened
with errors. Therefore, it is imperative that the errors
do not affect the conclusions and, as a result, the decisions made. Because decisions made on the basis of incorrect
numbers are too expensive. In theory, it’s easy to calculate the conversion,
but in practice, analytics systems often show numbers that are divorced from reality. How does Google Analytics determine conversion? He calculates the number of sessions with
the goal, divided by the total number of sessions. What is the problem? The fact is that the number of sessions per
user is a random parameter. Moreover, the error jumps not only by sources,
but also by time. The next problem is that GA records sessions
in which at least one event occurs. The fact is that on average one user performs
the target action from 1.1 to 2 times. It turns out that the average number of conversions
per user is also a random parameter. As a result, the application falls into CRM,
but is not recorded in GA. What does this lead to? Moreover, the number of leads in GA does not
agree with the number of leads in CRM. The fact is that the number of events does
not equal the number of sessions, and the number of sessions does not equal the number
of users. What is the problem? The fact is that the site does not order and
buy sessions, but people. What problems does this cause additionally? Due to the fact that GA does not record all
events, it is impossible to determine by which advertising channels they need to be distributed. As a result, it becomes impossible to calculate
the ROI of the channels. Marketing directors, to determine which campaigns
lead paying customers, bother with attribution models. What is an attribution model? This is a set of rules that determine the
principle of the distribution of value to the sources that generate customers. For example, a GA relies on a last click attribution
model. When revenue is attributed to the last ad
channel from which the request came. How does last click work? For example, a user accessed the site through
SEO, on the same day we returned it from the context, then he switched from the newsletter
and bought it. What will happen in this example? Email channel, understates contextual revenue
and CEO. For this reason, the intermediate channels
often do not meet the ROI, therefore, they seem unprofitable, which means that they must
either be optimized or disabled. How does the First click attribution model
work? In this model, revenue is counted towards
the channel from which the first transition was. For example, a user came from organics and
did not buy, but after a week, we squeezed it with a YAN retarget, and it turned. But 1st click revenue will be assigned to
the CEO channel. Income from the context will again be stolen. This ad channel has converted a customer,
but his ROI will again be underestimated. The next attribution model is Position based. A model in which 40% of the revenue is assigned
to the first channel, 40% to the last and 20% is distributed between the intermediate
ones. What is the limitation of this model? This model underestimates the ROI of channels
that attract users and overestimates the contribution of retarget, email newsletters, web pushing
and other tools that are involved in warming up users. The next model is Funnel Based. How does she work? She assigns weight to each source and distributes
income according to the weight model. How is conversion attributed in this model? According to the probability of the user passing
through the stages of the funnel. Those. conversion is credited to the source based
on probability. In essence, this method looks like a black
box. In addition, no matter how this model conjures
the weight of each channel, it still deducts revenue from paid RKs, which steadily underestimates
their ROI. In fact, we do not need an attribution model,
it is important for us to understand what to do with advertising campaigns. The attribution model is only needed to understand
which advertising campaigns generate revenue and which generate loss. Based on the attribution model, in fact, we
are looking for the answer to a simple question of what to do with paid channels in order
to highlight campaigns that need to be disabled, optimized, scaled, or not tormented and left
alone. Alternative attribution models, such as Post
click, also exist. How does this exotic work? If more than one paid channel is involved
in a multi-channel sequence, then post click assigns revenue to each campaign that was
in the chain but did not have the last click. For free channels, in the post click report,
only those leads are attributed for which the channel was the last in the chain, i.e.
when the number of conversions by post click=the number of conversions by last click. What does the post click model give us? She identifies campaigns that are definitely
necessary, either modify or disable. Thus, we get the maximum possible income from
the advertising channel in the conversions of which he participated. The next filed model is Indirect free, indirect
free revenue. How does this model work? It attributes income, if a paid campaign occurs
in the user’s path to the application, then in this case, the rhubarb is assigned to the
last paid channel. In general, the task of the marketer is not
to tinker with attribution models, but to attract the maximum number of leads for minimal
money. In essence, the task is to make one of 4 decisions:
turn off the campaign, optimize, scale, or leave it alone. But most marketers in their activities use
averaged data. In a typical approach, expenses are unloaded
from advertising offices, the number of applications from Google Analytics, and the number of clients
and transactions from CRM. After calculating the ROI channel by channel,
he concludes that, for example, direct and adwords are unprofitable, and the fb generates
income. But the real picture is often quite different. The fact is that different channels have different
conversion rates and different margins. What does such averaging lead to? Underestimation of successful campaigns and
overestimation of frankly weak indicators. As a result, due to such averaging, successful
advertising channels look worse. The marketer sees less profit and does not
understand how to invest in successful channels. Unprofitable channels, on the contrary, may
look advantageous, because their low margins are overstated by more successful ones. Here comes to the rescue, so punchy everyone
– pass-through analytics. What is cross-cutting analytics? This is a method of analyzing the effectiveness
of advertising campaigns, based on sales data, by tracking each customer through the entire
funnel. What is needed for analytics to become cross-cutting? It is necessary to identify the user at each
stage of the funnel. How is a user defined? By unique identifier clientID. When can the clientID not be written? Firstly, ad blockers, also when third-party
applications do not pass the cid value, when the site code in the browser is buggy, when
JS scripts are hung on the site, like on a Christmas tree, which conflict with each other
and the analytics cannot work out traffic give. When a user submits an application, he becomes
a lead. CRMs identify lead through userID, and payment
through transactionID. But CRM does not store ClientID, and Analytics
does not store userID. Therefore, in order for analytics to become
end-to-end, it is necessary to connect clientID, userID and transactionID. On the one hand, entrepreneurs want to make
decisions based on data, but on the other hand, they do not know what data they need
to collect, and most importantly, how to apply this data in practice. On the one hand, such a tool as end-to-end
analytics helps us work with advertising channels, but on the other hand, in our business, not
all tasks are related to advertising. Therefore, the correct performance-marketing,
without fail, relies on the unit economy. What is unit economics? This is an approach where all the decisions
we make are based on data. Thanks to this, we can evaluate how efficiently
our business functions. So that it doesn’t happen that we scaled
the business, but got a loss scaling. What does unit economics show? It shows how our business earns from the flow
of users. At the same time, it is important to take
into account the fact that money happens both with a plus sign and with a minus sign. When we launch a startup, we are at zero. Then, we start investing borrowed funds or
our own money, or we invest our forces, time and nerves, or all together. And there comes a moment when we begin to
earn as much as we spend and go through the so-called break-even point, but we are interested
in exactly the way to the break-even point, which is called the valley of death. It is the indicator of marginal profit from
the flow of users that shows us whether we can overcome this turning point. If this indicator is positive, i.e. when we
get more money from customers than we spend on attracting them, then our investments begin
to deflect. In fact, there are three approaches to the
use of unit economics. The first approach is when we are a startup,
when there is nothing but a crazy idea and we want to understand whether this business
model is worth implementing. Our primary goal is to determine which metrics
we need to achieve. Where do we answer our questions, at what
flow of users, at what conversion, at what retention and at what average check, we can
earn. The second approach is to understand the limit
values of metrics. What does this mean? When we try to understand which one, in principle,
we have a limit for each metric. The third approach is when we intend to scale. In this case, we begin to look for bottlenecks
in our marketing. When we have the current state of our metrics
and we want to grow, for example, 10 times. What does that require? Unit economics, in addition to searching for
optimal metric values, allows you to determine the path to change these metrics. As a result, the various characteristics of
our business must take the necessary values. Actually, unit economics allows us to connect
our business KPIs with metrics, where the mathematical model will help us find those
places that are critical for our business. Therefore, applying Goldratt’s theory of
constraints, taking into account the limitations of market conditions, we can evaluate the
economic efficiency of our activities and find the optimal metric configuration for
our business model. The next step in this process is to find a
bottleneck. What is a bottleneck? This is the point of application of force,
which allows you to get multiple growth. Why is it important to focus on bottlenecks? The fact is that if we begin to expand the
pipe not in a bottleneck, the system performance will not change. According to Goldratt’s theory of constraints,
business performance equals the performance of the bottleneck. If you imagine a business as a set of pipes
of different diameters, and take pipe sections under the business productivity, then the
exhaust of the entire system will be equal to the throughput of the bottleneck. Goldratt’s theory of constraints is a pretty
simple thing in terms of bottleneck optimization. When we find a bottleneck and perform the
necessary procedures to expand this bottleneck, then move on to the next bottleneck. As a result, we have such a process of continuous
improvement. What do we get? With the help of unit economics, we get a
chain of bottlenecks, as well as an understanding of how much and in what sequence we need to
change our metrics to improve our performance. What are we doing? We take our current unit economy, find the
optimal unit economy and determine the limits for each metric, under our optimal unit economy,
form a step-by-step algorithm for its changes. What does this algorithm give us? It allows us to determine what tasks, on which
metric and on which delta by metrics, we need to work on. Why is this necessary? This allows us to get the best way to achieve
goals in a minimum number of iterations. In fact, unit economics is used in two ways:
in taking metrics and in forecasting metrics. In the removal of metrics, everything is quite
simple, but when we try to predict the economy, in this case a number of complications arise. What is the reason for this? This is due to the fact that we need to predict
some parameters based on the values of other parameters. But around what is unit economics worth? The entire unit economy is built around the
user. What do we mean by user? The user is our basic essence, the user is
a person who, through advertising, got acquainted with our offer. The next entity that we operate in unit economics
is user acquisition, the number of users involved. What does this value show us? It shows how many users through marketing
with our site have contacted. Also, in the unit economy, we have such an
indicator as CPA, the cost of attracting a user. The next metric is LTV income from the client,
for the whole time of his life. The next metric is acquisition cost, marketing
budget. The next revenue metric is turnover. The next indicator is customer acquisition
cost. Zak takes into account all the costs we incur
in attracting a client. If we subtract from LTV – CAC, we get dirty
profit, from which we will contain our business. But for some reason, such a stereotype has
entrenched in the market that the ratio of LTV and CAC should be fulfilled as 3: 1. What is the trick? The fact is that the main criterion is whether
we can even scale our sales? If we can, then, in fact, the ratio itself
is not critical. It is important that the difference is positive. The key is to make money and understand how
we make money, and also to understand whether we can increase our sales volume. The next option is the conversion. In the unit economics, we are interested not
just in some kind of conversion, but in particular the conversion, which determines the process
of transition of a user to a client who has not bought from us yet. The next metric Average Price is the average
check. The average check entity is quite specific. How is the average bill calculated? This is a calculated value, namely, the number
of goods in the basket, multiplied by the average cost of the goods. The fact is that when we don’t have a monoproduct,
we can use such things as upsale and cosale. As a result, we have two parameters, with
which we can control our average check. Either by increasing the number of products
in the basket, or by the cost of each product separately. The next COGS metric is sales costs. COGS is a specific indicator that coincides
with a concept such as cost. Under the costs of securing sales, in the
unit economy we mean the costs without which the fact of the sale itself is impossible. The fact is, in order to sell a product, we
first need to bear the costs of purchasing this product, shipping and receiving money. The costs associated with these procedures
precisely determine what the costs of securing sales are. The next metric is First sale COGS. Additional costs that we bear on the very
first sale. An example of such costs may be, for example,
the cost of paying a commission to sellers. The next metric is the number of transactions
per client. This indicator is influenced by several parameters,
namely, conversion, percentage of user return, conversion to the second purchase and client’s
lifetime. What is the problem with this parameter? The fact is that APC marketers are often frivolous
and round off this value. Why is this indicator so important? The fact is that many businesses do not earn
on the first transaction. Therefore, it is extremely important to understand
with which transaction we will begin to go plus. The next indicator is Average Revenue per
Customer. ARPC is the average revenue per customer. It shows how much we earn from sales made
by the client for the selected period, excluding marketing costs. How is this indicator calculated? When a client comes to our business, he pays
us the average AvP check and from this amount we subtract the COGS costs that we incur to
secure the sale. In fact, this difference shows our profitability,
which we have from one transaction. But the client during his lifetime can make
a number of transactions, so we multiply this difference by APC and get the amount of money
that the client brings to us for the entire time of his life. However, in a business, there is often an
additional cost to a first sale. Accordingly, from all this term, we subtract
these additional costs for the first sale of 1sCOGS. As a result, we can begin to make the minimum
decision conclusions, namely, if we compare the income that the client brings us with
the costs of obtaining this client, then knowing that we have more ARPC than CAC, we will understand
that our business steady. This difference is crucial for scaling. Because when scaling a business, by some metrics,
we can begin to sag. Therefore, it is always useful to have an
extra margin of safety in reserve. The following metric is Average Revenue per
User, average revenue per user, excluding marketing costs. ARPU is an important measure for evaluating
business performance. Comparing it with CPA, we can evaluate the
return on investment. And the key indicator of the unit economy,
for which all this was conceived, is the Contribution Margin, the marginal profit from the flow
of users. This is a value that shows how effectively
our online marketing functions. What does Contribution Margin show? Contribution Margin compares the cost of getting
one user with the revenue received from this user. As a result, we can evaluate how good our
economy is. In fact, the entire data-driven decision approach
comes down to comparing the Contribution Margin from decision to decision. What does it look like? For example, we want to earn a million in
our business by a certain date. After that, we decompose our million into
metrics, namely, what kind of user flow we need to provide, how much we need to pay for
each user, how much we can take money from each client, what kind of customer return
we have and based on these values, we calculate Contribution Margin
value. What are we doing? We take the task pool and rank them according
to the changes in the Contribution Margin value. The method is very simple, i.e. those values
that will give us a gain of Contribution Margin are correct,
those that do not give a gain are incorrect. Based on the goals, we collect the configuration
of our metrics – this is the number of users, the conversion from users to customers, the
average check and the costs that we incur on each sale. Based on this, we can digitize the effectiveness
of our sales process. After that, we estimate that we must change
in these metrics in order to reach the necessary financial indicators. Therefore, it is so important for us to understand
what metrics we need to change, and most importantly how much they need to be changed. This is the key essence of unit economics,
namely, unit economics allows us to evaluate and rank tasks. In essence, to build a decision-making system. Thanks to the unit economy, we can determine
which strategy for reaching the desired indicators is optimal for our business. On this optimistic note, I propose to round
off this video. Therefore, I want to wish that the unit economy
in your business always converges. Thank you for watching, subscribe to the channel,
click the bell so as not to miss the next issue, express your point of view in the comments,
like and see you in the next video. Goodbye!

6 thoughts on “✅ Performance-маркетинг. Юнит-экономика

  1. Performance-маркетинг. Юнит-экономика – комментируйте, высказывайте свою точку зрения и подписывайтесь на канал → https://goo.gl/yavJVw

  2. А можете показывать пример настройки на видеоряде? Это было бы очень круто!!

  3. Благодаря вашему видео нашел необычный способ привлечения лидов, это и правда перфоманс)

  4. Extraordinary Work, I enjoyed it a lot!, Check My New Rock/Metal Album 'Monish Jasbird – Death Blow' available on all digital stores & streaming services, i guarantee it will be your favourite, links available on my channel, you might like 🙂

  5. Информативное видео, очень понятно разобрали тему Юнит-экономики

Leave comment

Your email address will not be published. Required fields are marked with *.