I typically run a couple of a hundred queries a week around SEO long tails to try and capture SEO posts that may not be on my radar, but worth a read. The nature of these change from time to time based on my current interests – after all, I started out as a Small Business SEO and now work at advising Big Brands on SEO. As may be apparent, due to changing interests, many of these queries are around Big Business (or Enterprise) level key phrases. One of these being around budgeting and estimating Return on SEO Investment (aka SEO ROI).

Surprisingly, there aren’t many resources available to businesses to realistically budget their SEO spend, for many reasons, one of the bigger ones in my opinion the difficulty in calculating life time value of a position. After all, it’s not like SEO rankings are like PPC, where every click has a direct attributable and potentially a onetime cost (of coursed based solely on a last click metric. Under other models such as a weighted average model the PPC click has a higher lifetime value than the last click model). So imagine my surprise that that one of these keywords “SEO Budgeting” led me to an eHow article. Now for those of you who may not be familiar with the content farm nature of eHow, you may want to visit my post on content farms.

The article (which I refuse to link to) was written by a company that ,surprise surprise, offers Search Marketing services. Now I don’t want to blast that company – what they do is none of my business. But the information they supply to the general public is. So with the hope that SEOmoz ranks higher for this keyphrase, I would like to dissect their advice and point out some better resources for the SEO planner. 

Difficulty:

The level of difficulty indicated is Easy. I wouldn’t say it’s easy to decide how much money needs to be spent on SEO. There are so many assumptions that need to be made – take for example some of the top line questions you should be asking:

• What is your sites SEO Swot Analysis?
• What are you trying to rank for?
• What is your average return on your current SEO?
• What level of budget do you have available?

I would give difficulty level between Moderate and Difficult, depending on the size of the business to the breadth of keywords that the site is trying to rank for.  

Ways to Gauge Budget:

Once again, in my view this advice fails. The three things that have been advised are:

• base your campaign on a past successful campaign. If you had a past successful campaign, you probably not going to be looking for SEO budgeting advice, considering that you have already run successful SEO.
• “designate a target percentage of total sales you wish to achieve though SEO marketing and work backwards to make this figure a reality.” This is a feasible way to predict future budgeting ONLY if you have run a campaign previously. Ideally you should base your budgets on the opportunity available with reference to your available spend and ROI.

What portion of your budget should SEO be made up off:

Sorry, but that is utter rubbish. Where did those figures come from? Very poor advice. SEO, like any other marketing strategy should be run by either branding decisions where you work with maximum exposure possible) or on measurable ROI.

To explain that a bit further – take for example you have an opportunity to rank for a pot of 4 keywords:

If you had to choose, you would probably go for Green and Yellow widgets, if traffic was your goal right?   Now look at the details below:

A Bit more depth – you have evaluated the cost of acquiring links, writing content etc for those specific keywords, and worked out the cost of ranking that keyword. Now if you had the full £9,500, you would probably go for all those keywords. However, budgets are finite, so you may not have the full £9,500 to spend. So what else do you need?   You need to work out the revenue on those keywords:

Remember, every keyword has its own revenue potential.  Just because a keyword has high volume, doesn’t mean that its value in proportion will be the same as a low volume keyword, in fact, the returns on volume are normally diminishing. Which is why you may want to add one more metric, ROI:

Isn’t that an interesting metric? The keyword with the highest volume has the lowest return, while the third lowest keyword has the highest return. Taking your keyword evaluations this far help you nail down the most profitable keywords that you should be aiming for, which in return gives you a view of getting the right budgets put together.

OK so the big questions are HOW do I work out those figures above?

To start with, you need to work out potential traffic, by estimating it.  Kate has done a brilliant job of showing you how to predict your traffic. I dont think I need to reguritate those examples here :). 

The second step is to work out conversion / return – now there isn’t an easy way to do this, but you have three routes – either grab your current conversion by “generic” keywords, and use an estimation (this will probably NOT give you’re the variant ROI’s as in my example above) or segment those generic keywords into behavior patterns. By this I mean you could breakdown your current generic keywords into the type of keyword, and then break that down a step further into the length of the query – rule of thumb experience says that the longer the tail, the closer the user is to the purchase journey. This means you could apply a higher average conversion rate to longer tail – but using your data segmented into keyword type and then into tail, will allow you to get a rough estimate how that mechanic works currently for our site. You can use those figures to estimate the sales.

However the two examples above require you to have some history. What would you do if you didn’t have any history? I would use an isolated PPC test to judge the conversion value of any pot of keywords.  The isolated tests will probably provide you with usable variants of values that you can use for conversion.  Rand covered this idea way back as part of his Head Smacking tips - Using PPC as an indicator of Success. 

How do you work out the cost of ranking for any single keyword? This is probably one of the more difficult questions to answer. However, some of the key indicators of cost of any given keyword can be derived from working backwards from your Competitor audits. I think the biggest variable outlay for any SEO ranking is the cost of acquiring anchor based text links for the given keyword. The second variable is content rewrite or content enhancement, which probably is lower than the link cost. To these variable costs, I would proportion other SEO “fixed” costs such as any related fees, overall other developments impacting SEO, time of staff spent on SEO improvements etc. If anyone has any real good ideas to estimate the cost of ranking, please share!

Summary

Unfortunately, if you kept up with my ramblings so far, you will notice that there isn’t a fixed art of estimating the budget, but there are some arbitrary estimates that you can use to work out the portioning of spend to SEO. The key thing in my opinion, is to invest in ROI while staying within your allocated budgets.  

Disclaimer

This is my methodology or route to working our SEO Budgets, please make sure that whatever decision you make is best catered for your individual example. Also, the examples given are very simple views of setting keyword base targets - carrying out this exercise for 1000's of keywords isnt exactly easy, but it is worth the time and effort.

Resources:  

• Jamie wrote a very useful post on convincing your boss to spend money on CRO. He includes a link to a very useful spreadsheet, which you can easily adapt to work out SEO ROI instead of CRO ROI.
• This is a very useful PPC ROI calculator, which yet again be adapted to work out your SEO ROI figures…
• Gab specializes in SEO ROI, and he gives us 8 steps to forecast ROI from SEO.
• The SEO Slingshot blog has a two part article on evaluating the Value of SEO
• Last but not least, Aaron’s SEOBook blog looks at the Value of Google Rankings in some serious Macroeconomic detail.