Conversion Rates: Rule of Thumb?

[jamesa]

I know this is nearly impossible to answer, but I could really use some guidance here.

I'm putting together a campaign and one of the factors in determining the overall cost and ROI is, of course, the conversion rate. I have no prior data but I need to plug in a conversion rate that's reasonably realistic just to work out the math; the feasibilty.

Initially I was thinking in the 1-3% range. Is that typically low, average or high? One campaign I'm considering breaks even at around 3% and profits at nearly 6%. Is that too ambitious?

So if you were in my shoes what numbers would you plug in? What's a realistic range for you (low, high, average)?

[Shak]

what type of shoes are you wearing, what size are they and do you have socks on, and maybe your feet smell :)

On a serious note, totally impossible, conversion rate partly depends on traffic, but other factors such as:

design
pricing
customer service
usability etc

sorry cant be more helpful

Shak

[onlineleben]

> partly depends on traffic <
and there it depends on how targeted your traffic is.

[eWhisper]

I've had to make estimates based on insufficient data before for CEOs. I always estimate on the lowest side possible, and tell them that it's an estimate based on total speculation and means nothing, but here's the report you wanted. This way, if the campaign doesn't work, I can tell them it was speculation, and if it does work, it often has a much higher ROI than predicted which makes them real happy.

However, they need to realize that the report truly means nothing without actual data - and if they can't understand that, ask them to run it by the company's accountant, as they hate working with anything but actual numbers.

[qfguy]

To me the importance of conversion rate has more to do with margin than anything else.

If I make something that costs me $1 but I sell it for $100 I would be willing to suffer a much more terrible conversion rate than if I can only sell it for $1.01.

[bigjohnt]

a huge number of variables go into an analysis like this.

I simplify, and rely on an old direct marketing formula of 1 - 2% response for forecasting.

One extremely competitive client has a fairly good 1.8% response (lead)rate, but converts approximately 35% of those leads, sooo, we have a totally new "effective conversion" rate of 35% of the original 1.8%, or what appears to be a dismal .63%, but which is in reality the best they've ever done when measuring cost per sale. :)

The keys are relevance, competitive cost, and available inventory of searches. This assumes the site is very well designed for action - making it easy to do business. (Not to impress visitors with the technical expertise of the design and IT staff.)

To confuse the issue, I've found that the source of the traffic skews things also.. even the same search terms on different engines varies wildly, and week to week, the same search term/engine combination fluctuates.

For example, during the week with the lowest number of visitors, we had the highest conversion rate (over 8%) and a conversion rate of over 90% from those leads. This would indicate a highly motivated visitor during that period.

During the week with the highest number of visitors, we had a low visit to action rate of less than .7% - and conversion of less than 30%. However the huge number of visitors netted a larger number of sales for the week.

After literally hundreds of hours of number crunching, I've found no absolute formula or totally accurate guideline.

Best to test a bunch of traffic from various sources, then narrow it down and use a historical figure, but for forecasting, I stand behind 1 - 2% expectation. If its better, great. I love to underpromise and overdeliver!

[shorebreak]

Great data bigjohnt, wish there were more like you out there.

[chewy]

It really is only a number.

It gets much interesting if you can say it was X% per month on average and it is now Y% per month on average.

It is not the number that matters - it is the delta.

We've taken a 3% conversion rate and turned it into 6%.

That is a delta of 100% and we see it month after month, some better, some worse.

Sure, I know they want a number - and they want to know how it compares with others.

I tell em it really only matters in comparison with itself over time. (remember same store sales?)

ROI is a bit more useful, and even better over time, as long as the assumptions,equations, etc are equal, the data is gathered correctly, etc.

C.

[jamesa]

Thanks everyone. Especially bigjohnt, that's exactly the kid of info I was looking for.

So maybe the question should be how do you determine your initial bid amount? Obviously that gets revised once the campaign gets going, but what do you factor in when starting out? Ya'll have to be making some assumptions here. :)

[skibum]

I think that depends on the funds available and how fast you want to learn. If its something completely new and there is some testing budget I'd shoot for the top 4 spots across the board to start out, collect some data and then have a somewhat better idea of what works and what doesn't.

It's the price to pay to be able to be able to do somewhat better forecasting. Like bigjohnt mentioned there is not always (usually?) consistency. Seasonality plays into most any business, it seems that some keywords "wear out" after a while, sometimes its the creative or both.

[eWhisper]

Ya'll have to be making some assumptions here.

To make an initial assumption, with a company that has no idea about their ROI, I look at the bids to see if there seems to be a bidding war that has pushed the bids higher than is reasonable, if yes, then we look for a good niche just below the warring parties - and see if we can price pin them to bring the bids down.

If the bids seem fairly consistant, then we try to jump into the first 3 spots on OV, or the 2 premium positions on AdWords.

The first couple weeks, we try not to spend a huge amount of money as we're tracking closely on a daily basis their ROI, and as it's easy to be off, they could lose or make money the first month. Although at worst it's usually break even. We generally assume a 1% conversion rate, which is generally low for niche phrases, and maybe high or low for broad phrases. After we now have some good historical data, we can easily figure out where to keep their bids and how to get them the best ROI possible.