• by David Rodnitzky , Op-Ed Contributor, June 9, 2014

SEM agencies can be pricey. The standard fee is based on the “percentage of spend” model, meaning the more you spend on SEM, the more the agency gets paid. So if your agency charges you 10% and you spend $300,000 a month, your monthly bill is $30,000. That’s a lot of dough! So how do you decide whether that cost is worth it? I’ve tried to create a mathematical formula to help you solve this age-old question.

To get to the correct answer, you’ll first need to know a few pieces of information about your company’s finances:

1. What are the fixed costs associated with your internal team? I usually think of this as your SEM team’s salaries + 10% for benefits and office space.

2. If you hired an SEM agency, how many fewer people could you employ on your internal team? Note this does not mean that you have to fire people when you get an external agency. Many companies choose to hire an external agency in lieu of hiring new internal team members, though some companies do decide to eliminate internal positions once an agency is in place.

3. What is your current gross profit per one dollar of SEM spend? This calculation will vary greatly depending on the type of business you run. For example, an ecommerce company might determine gross profit by looking at revenue minus cost of goods sold (COGS – how much you have to pay for the item you are selling). A service provider would determine gross profit by revenue per contract minus services costs.

4. How much will your SEM agency charge you for their services? Make sure to include all costs: agency fees, technology fees, creative fees, etc.

If you can answer these four questions, you are on your way to calculating the value of an SEM agency. What we need to do first is calculate the net cost of the SEM agency. To do this, we use this equation:

(SEM Agency Fees) - (Before Agency Fixed Costs – After Agency Fixed Costs) = Net SEM Agency Cost

Let’s say that we’ve determined that we were paying our internal SEM team $8,000/mo. prior to working with an agency, but we will only pay $5,000 a month after. And let’s say that the agency is going to cost us $13,000/mo. Using our equation, the net cost of the agency is $10,000 ($13,000 – ($8,000-$5,000)).

Now that we know that the agency is costing us an additional $10,000 per month, we can figure out how much additional gross profit per dollar or additional spend at the current gross profit per dollar the agency needs to drive to justify its costs.

Let’s say, for example, that we are currently spending $200,000/month on SEM and our average gross profit per dollar is $.25. If we want to continue spending the same amount of money, to justify the net $10K in additional fees, the agency needs to increase our gross profit per dollar to $.315 (or by $.065). The equation to discover this amount is here:

($200,000 times X) – $13,000 = ($200,000 X .25)

Alternatively, if we are OK with the agency increasing spend but not increasing the gross profit per dollar, the equation would be:

(X times $.25) - $13,000 = ($200,000 X .25)

Which gets us to a spend of $252,000 per month.

There are, of course, variations on this theme where you can play with increases in both spend and gross profit per dollar. There are also instances where the external agency might save you money over the costs of your internal team, meaning that the agency could actually decrease your performance and still be a worthwhile investment.

It’s also important to recognize that agencies need time to ramp up on accounts, which often means that performance in the initial months of a relationship will not be as strong as it is further down the road, once the agency has gotten through a few rounds of testing. So you might want to hedge your bets a bit by assuming that the agency does not drive incremental performance right away.

To be clear, I don’t suggest that you base a decision like this solely on the direct financial impact. There are a lot of non-mathematical reasons to consider when weighing in-house versus external options. Creating a formula to sanity-check your decision -- and measure the results afterwards -- at least removes some subjectivity. Given the analytical nature of most SEMs, I suspect this idea will be welcomed with open arms -- and meticulous pivot tables!