An e-commerce's budget definition and allocation can be the reason why it succeeds and sky rockets... Or the reason it goes bankrupt.
We all know marketing is a crucial part of an e-commerce's success. If people don't know what you're selling, why you're better than your competitors, or if they don't even know you and your products exists, they won't buy from you.
Of course, a huge part of your marketing's success are the actual marketing campaigns: the creatives, the message, the colors... But that's not it. How much money you spend on each channel can mean the difference between a marketing spend higher than the revenue it brings, or a super profitable ROAS margin. This is mainly affected by two things: how much money you spend in each channel, and how you divide you budget between campaigns inside any of your channels.
Optimization inside a channel includes how many campaigns you have active, how much money you're spending in each one, the bidding you set, the creatives you choose, the message promoted, the type of ad shown, the audience targeted, all the daily adjustments you make to each one of those fields (at least for the digital campaigns), etc.
Optimization between channels means defining how much money you spend in which channel, and consequently, how much money you spend in advertising in general.
The digital miracle
Over the last few years, lots of new advertising channels have emerged, opening up exciting new possibilities for advertisers. Digital marketing means you can turn on, turn off and modify any campaign you want at any time of the day. So if you realize there was a mistake, you can easily solve it without millions of people seeing it, saving you that money. You can also set up random promotions, or make adjustments if a shipment is delayed, or if you're simply running low on budget.
Not only this: digital marketing also means loads of new ad types: Facebook's carousel, Google's search ads, YouTube video ads, Taboola's articles, Criteo's banners... However we find ourselves with the 'paradox of choice': there are all these new channel and ad types, and countless ways to distribute our marketing budget... Which one will be the right one?
Fortunately, the new digital channels are much easier to analyze in terms of efficiency. Whereas with a street ad you don't have an actual number of how many people passed by it, how many people actually saw it (and didn't just pass by looking at their phones), and how many's attention you managed to catch; with digital ads, this becomes an easy task. Every digital marketing channel will tell you exactly how many impressions your ad had, some of them will tell you how many of them were 'viewable impressions', and they'll all tell you how many clicks it had (how many people's initial attention you managed to get a hold on). Furthermore, if someone goes ahead and buy something, you'll know that sale was driven through that ad.
Awesome! Then it's really easy: you just put your money in the sales-driving ads/channels, and cut the spend from the rest, right? Wrong. Turns out people don't usually buy immediately after seeing an ad: we do a bit more research, go back home, think it over, then maybe come across another ad, and think a bit more about it... and eventually we buy.
The digital nightmare
This natural shopping behavior of ours raises 2 main complications:
a) Attribution. Users pass through a multiple number of sources since they first learn about your existence until they make their final purchase. So how do you decide which was the decisive interaction? Was it the first one, because no awareness means no purchase? Was it the last one, because that's when the user actually bought something? Come to think of it, maybe if she hadn't had all the other middle interactions, she might have forgotten about you and never bought, so those are important as well... So how do you weigh them?
b) Losing track. This is more commonly referred to as cross-device. Basically, it's really easy for you to lose track of a user in the middle of all different devices he uses - work computer, personal laptop, at least one mobile phone, iPad... How do you know that John is the same person who browsed your site two days ago from his smartphone after seeing an ad on Facebook, and the one who made a purchase on a Mac this morning? The only way would be if there was something to connect both interactions, such as an active session for some platform in both devices. For example: he is logged in to Facebook both on his phone and in his laptop. However, this means only Facebook knows that it's the same person, and it will attribute the sale to its ad. But what if John is also on Gmail in both devices, and he happened to also look for your page after seeing your ad? Google would also attribute this sale to its search. This is not an attribution problem: the actual problem is, Facebook and Google don't talk to each other. They don't know John had the other interaction. And neither do you. So when you see a customer's journey to a purchase, you are likely to be missing big chunks of information. More information on customer journeys can be found here.
So how do you define you budget and how do you allocate it?
Budget definition & allocation
Defining and optimizing your budget can be an imposing task because of its importance and delicacy.
The first thing to do is look at your numbers with a critical eye, and analyze how every change you make affects you campaigns' interaction and sales.
We think the basic steps for every budget allocation are:
- Take a look at you customers' journey to a purchase. See not only the channels which are bringing you sales, but all the ones who are helping you get those sales (all the steps of the way before the final purchase)
- Target the right audience. This means: segment your audience in every way you can. Show watches to people who like watches, and show books to people who like books. Platforms want to improve user experience, so they'll prefer to show high-relevant content to their browsers. This preference is not only reflected in how many times they show your ad, but also how much you pay them to show it. They make it harder and more expensive for you to show content of lower relevance.
- Treat different users, differently. If someone has already come to your page, they (should) already know what it is you sell. These people need a different message than people who don't know you: there comes a point where only seeing you brand, they'll make the association with the kinds of products you sell, so you don't necessarily need to give them that introduction: a picture of the product and you brand will be enough. Turns out, these people who already know you are much cheaper to target, and it gets cheaper and cheaper depending on the level of interaction they've had with you. For starters, they already know you. If they've also given you their email, you can target them for free with a newsletter. And if they have already from you, it means they have tried you -and we'd expect like you. So it's going to take much less effort to convince them to buy a again, than to convince someone who doesn't know you that you might be good, look at your products, trust you with their card details, and expect a good product in return.
- Test everything. Don't assume you know 'what works' or 'what your customer want'. Innovate: try new creatives, messages, promotions, colors, audiences, ad types, biddings, marketing channels... You'll surely be surprised with at least one of the results!
- Optimize, optimize, optimize. Don't make a campaign and forget about it until the end of the month. As we mentioned, the ability to change instantly the configuration of a campaign is something unparalleled in traditional marketing. Make it your ally! Take a close look at numbers, change audiences when they are saturating, refresh your ads, take advantage of local events or festivities. This goes for moving budget both between campaigns inside a channel,
How do we do it?
As there are so many variables, and so many relations we might not be aware of, at Sensai we think the answer for an optimal budget allocation lies in a combination between man's expertise and machine's intelligence, which is why we have built a tool based in business analysis and know-how of great marketers, as well as machine learning algorithms.
If you want to know exactly how it works and why it's awesome, please read our next post!