Thanks again to everyone reading this and contributing to the success of this website. We have had an amazing first 75 days and would like to share what we have learnt, and we hope you will learn something new from what we learnt.
In part 1 of this series, we covered visitors and visitor interaction. You can read part 1 here.
In part 2 we cover our experience of advertising this website online. It has been quite a learning and very surprising.
One of the things you get for free when you start a new website is free online advertising credit. Most online advertising networks will give you free credit. And by most we mean Facebook, LinkedIn, Twitter (occasionally), Bing/Yahoo and Google (you must spend a little of your own money first in the case of Google – in the UK £25 of your money will be topped up by £75 of free credit from them).
As an experiment we created adverts for different networks, some of our sample adverts are as shown below.
Select your audience
All advertising networks let you select your audience. You can select specific geographical regions (ranging from entire countries to small towns to include or exclude), the kind of device being used (Desktop, tablet or smartphone) and demographics. The demographics is an interesting thing. In the case of some advertising networks (not difficult to guess those networks since the track everything about you) this is exhaustive, you can select people by likes/dislikes, type of opinion (brand perception about a given brand or support for political causes), the normal things like age, gender, how expensive their primary device is (from an expensive iPhone X to an inexpensive Android device), their level of education, the languages they speak, where they did their education, their estimated income and many more.
Bidding wars
Most networks will let you bid on certain keywords or interests. When people search for those keywords your adverts are shown. Other networks will let you select interests and people with those interests are shown adverts.
Decide the type of ad (and design it) and you are ready to go
We just designed simple text ads and were ready to go. But it didn’t work. Here’s why:
We first selected keywords and interests that would get us our target audience. Our keywords were mainly 1) Finance Blog 2) Financial Blog 3) Economics Blog 4) Business News 5) Markets Blog and a few more. Our target interests were those people reading 1) Financial Times 2) The Economist 3) CNBC and generally interested in 1) Business News 2) Finance News 3) Stock Market News.
We selected Canada, the US and the UK as our target locations. We were told we should start with a bid of £0.05 (we are UK based for those who are wondering). No adverts were shown, we were told to up our bid to £0.08 and add a few keywords and interests. Still no adverts shown. There is a constant bidding system, so the bids kept going up. Our first adverts were shown when we bid £0.32. Great, advertising is live now.
We initially had 4-5 clicks to our website a day. Those who reached our site spent an average of 2 seconds on the site and didn’t click any post/article.
We then asked the question – is it just us? Is our advertising not good or is our website not good enough for people to explore it? Well, it seems this is “normal” in the internet advertising world.
Anyway, since we had to spend the free credit we tried another thing. We bid the minimum (£0.03 in some cases) and selected the entire world as a geographical target. The result? We now had 20+ clicks to the website a day. On an average, each person spent 1 second on the site.
On further investigation we found that most our clicks for which we paid £0.03 to £0.08 came at around 4am local time from each country (India, Thailand, Indonesia, Philippines) and all were from Android devices with a resolution of 360 * 640. We then searched the internet to find that other people found exactly the same thing with their own advertising. So, let us get this right, people are searching for a financial blog at 4 am their local time on their mobile then reach our site and spend 1 second on it.
Further details from the internet revealed exactly this – you aren’t going to get much “genuine” traffic from internet advertising. Strangely a number of websites had traffic from advertising between 2 am and 5 am local time of their audience.
We again changed our strategy to target advertising audience in Canada, the US and the UK and did get some “genuine” traffic – people who read a few things on our site. Don’t know if they will return to the site but at least some success.
We aggregated some benchmark statistics from the internet. First some terminology explanation,
Search ads are those that appear when a potential customer searches for your product, service or a related item on a search engine such as Google, Bing or Yahoo.
Display ads appear on a website while browsing. They are typically banner ads located around the page with attention-grabbing headlines or images.
Click through rate (CTR) reveals how often people who view your ad end up actually clicking it.
Cost per Click (CPC) is the price you pay whenever someone clicks on your advert.
Conversion rate tracks the amount of conversions you get from a certain ad.
Cost per Action or Cost per Acquisition (CPA) is the amount you pay every time an action is performed involving an ad.
The average cost of an action is staggering. Some might argue it is more effective that other channels of advertising. For instance, we get loads of unsolicited post here in the UK for high interest low credit rating credit cards. Assuming 1 in 200 people who get that in their post actually gets one of those cards it must be costing them at least £150 to acquire that customer using that channel of advertising. The same thing can be achieved online at the same price if at least 1 in around 200 people who click that advert (assuming the cost per click is £0.75 for the credit card advert) actually apply for it.
Online advertising was ineffective for us and there was a feeling that most clicks were fake (note: we said “feeling”, we aren’t drawing conclusions here). It didn’t work for us but might work for bigger advertisers who have learned to make most of the system.