Obviously my service is not a compulsive buy, due to the cost. People need some time to think about this kind of purchase and may come back a few times before they buy. But how long?
Does anyone have experience with a more expensive service? Am I to expect a few week lag before orders start coming in?
I know this is general, but if there are people here who offer more expensive services or products, I'd like to hear about your start-up phase and how long it took before you could predict a pattern of sales.
Help is very much appreciated. Thank you.
There are many variables you have to check out and this forum has many great discussions that would be of help.
I use it more to "be present", to build up a brand image for our company. I can have my ads displayed even with keywords which only remotely have to do with our products and which therefore do not justify (at the moment) the time being spent on seperate webpages for these keywords.
When they started web advertising, they did not see a significent increase in sales for a little over a month. Their sales slowly grew for about 6 months until it leveled off at a fairly consistent number of increased sales.
They don't know the exact sales figures from AdWords (or any of their advertising for that matter), they just know that their total sales went up when they use it. If they stop using AdWords, they don't see a decrease in sales for 1-2 weeks, and a trickle effect for the next 1-3 months, so it's quite difficult to track.
They have gone with the model (and I don't endorse this at all - I like to track things better - its just their model)of we're spending x amount of our profits on advertising. Our web ads get 25% of our total ad budget - spend it where you think it'll do us the most good.
My suggestion, if you do not have a higher priority task, is to break up a "conversion" into a series of smaller baby steps that do not alienate your customers.
For example, one baby step might be adding a "newsletter" about your products / services. Now, instead of worrying about whether you are converting your leads into cash, worry about whether or not people are signing up for your newsletter.
If you are selling plasma TVs, have a small review newsletter on high end home entertainment devices.
With the newsletter use suitable informational / selling techniques that educate your users and get them interested in buying your products. If you get other products online, you can also upsell / cross sell them later.
Newsletters are not always appropiate and should not be shoved into your customers face. Other baby steps might be a software demo or a white paper.
Another babystep which doesn't require much in the way of change is just counting pagecounts. If certain keywords only lead to users leaving your website immediately and others lead to the user trying to find out lots of information about you by clicking around, I suggest you optimize on the latter.
I thought about using the conversion tracking tool from AdWords, but I'm hesitant since I do not really want to have the "Google Conversion Tracking" graphic on the site.
Any opinions regarding this graphic?
One of my clients sells a service that ranges from $25k and up. It often takes more than a week to know the conversion rate (and sometimes conversions appear 6 months later), and as many of them are phone calls, emails, and ongoing dialogues before its purchased, it's almost impossible to get accurate conversion stats for AdWords vs other advertising.
I suspect that the use of cookies will help, particularly in follow-up HTML emails that invite a click. We use 90 day cookies on the web site only and see conversions where the first visit was 3 months ago.
A side effect of continuous advertising is often a huge increase in direct referrals to the site on the scale of 100%+ per year and tons of sales and leads that cannot be directly tied to any advertising campaign.
In addition, with this type of long term data its possible to spot seasonality trends every year. An isolated six month campaign may look like a total flop as leads or sales decrease month over month while it is running, however, when compared to the same period the previous year it appears to be very successful.
The biggest challenge is often getting clients to stick with a campaign long enough to see the big picture and watch their direct referals in particular jump year over year along with sales revenue or leads generated that can't be directly tied to a campaign.
I suspect that the use of cookies will help, particularly in follow-up HTML emails that invite a click. We use 90 day cookies on the web site only and see conversions where the first visit was 3 months ago.
Cookies are used, special 800 phone numbers are used, etc. However, as all the sales are completed either via the phone or in person, potential customers end up with direct phone lines over the special 800 ones, etc.
Stats are great for traditional ecommerce. And I'm not endorsing people don't track (I've gone through **** getting this company to get web 800 numbers and start to do some tracking).
However, Skibum makes a great point. I think some companies can get buried in stats and ROI measuring.
There have been many effective companies built on YP advertising. They can't tell you which phone book every person who bought from them used (some places have 3-5 phone books - and the average consumer doesn't know which one they happened to look in to find you) and those companies advertise in 100s of diferent YPs. They can track zip areas to know which areas to invest in yps - but if one gave them a negative ROI - they still wouldn't stop using it - they would adjust their ad size/layout/etc.
An offline company, when advertising fails, often looks to changing how they approached the ad and demographic to correct the problem. Often online companies, when they don't see a positive return, dump the advertising as unprofitable instead of trying to fix it.
Very different ways of reacting to the same numbers.
What you can not measure you can not manage.
If you are smart and creative, I have yet to see a campaign you can not measure with very little work.
For example, even with TV advertising campaigns they will target certain cities first to see the affect on sales. However, in order to do this, they have to blanket that city with a campaign in order to be sure that their advertising has not just been a coincidental flux.
I don't think the question is - should I measure?
The question is - should I spend enough to reduce the probability that my measurements are erroneous.
Blanketing an area to see an increase in sales for that city is a form of measurement yes. But by the same token, as stated above, they are measuring in the same way by knowing that their sales go up and down by a some amount with or without certain types of advertsing.
As they only make a few sales a week, the difference of 5-10 sales could be a 25% change in total business. While 25% is a significent number, the number 5 is not an easily quantifible number as when you have 100s of ads, and on average, you only need 1 YP ad to convert once every 2 years for the ad to be profitable. Likewise, one AdWords conversion a month out of 10k+ visitors means they made money on AdWords.
How do you measure one single ad in a yellow page and know that that ad made one sale in 3 years?
Likewise, how do you measure the AdWords branding of seeing a company name when a visitor might have seen the ad, but then entered the website through a totally seperate medium after seeing the ad multiple times? The AdWords ad helped get that buyer, but it didn't actually bring that buyer to the site.
I generally run PPC programs using the old direct marketing rule of thumb: Run client acquisition at marginal-cost breakeven and you'll be on average very profitable.
Just because the AdWords tracker or whatever click-counter du jour is inaccurate does not mean it is impossible to measure what is going on.
Let me give you an example. I worked with a fortune 100 company which had an extremely impressive brand, however it was important to them that they were able to measure a certain marketing medium (banner ads) accurately.
So, basically, they created a brand relatively overnight to weed out the flux and used the banner ads to lead clicks to the product and measured the ROI. While the junior guys had all these reports for management, at the end of the day all they looked at was income in, advertising spend out.
They did this because they wanted to to be able to measure the impact accurately of these marketing campaigns. It was important that they simply didn't make ad buys for the sake of ad buys.
When discussing it is tempting to assume everything is black and white. I generally assume when writing that people will know that I mean a statistical analysis needs to be performed on everything and margins of error will always exist. Measuring is about being precise as possible, it is not about being exact.
