An estimate's an estimate, right?
Not necessarily. Agencies use a few different types of estimates, and each has it's positives and negatives. I've outlined a few of these common types of quotes below, with my take on each based on my experience.
Fixed bids have a logical simplicity—the company wants X, and it will cost Y. If there are significant changes along the way, the agency will submit a change order for approval. A huge limiting factor is, as discussed earlier, requirements are rarely clear before estimates have to be made as part of a proposal.
This leaves the agency to protect themselves with a significant buffer or play fast and loose and hope underbidding doesn’t bite them in the ass. Sometimes this is done in earnest, but other times it’s part of a sketchy strategy to first win the business and then do their best to extract more once the client is locked in. The real weakness of the fixed bid is most visible towards the end of a project when the incentives of both parties become diametrically opposed.
For the agency, every client request or delay is costing project profitability. They will push the production team toward cutting corners, quick turnaround, and getting it done.
For the client, they are incentivized to keep pushing until everything is perfect, even if it means redoing already approved work as their costs are fixed.
At S4, we moved away from fixed bids after a couple years due to getting badly burned a few times. You don’t see this as a primary model for agencies except for ones that focus on government RFP work. That being said, at S4 we have introduced a product called Continuous Improvement that focuses on launching a new site quickly followed immediately by iterative sprints. In this case, the cost of site development is fixed because the scope is explicitly flexible, as we push non-critical tasks to a backlog that will be pulled from following launch. This is discussed in more detail later.
Time and materials is a common approach for many agencies. Like a lawyer, you’re paying for the time a group of professionals works on your behalf. Some agencies use rates based on the role or experience of the person doing the work. Other agencies have different rates for the type of work being done (i.e. design versus backend development).
The downside is obvious to prospective clients—you’re on the hook for the time you use even if you don’t get the output you expected. The downside to agencies is likewise obvious—you can lose money if you end up with a demanding client.
Likewise, the incentives are lopsided—a client wants work performed as efficiently as possible by people or roles that have the lowest rate who can perform the job. For agencies, efficiency is less of a concern compared to quality. Employees at the agency may have billable hour quotas and be inclined to bill first and ask questions later.
At the same time, a good agency knows that unexpected bills or overages need to be communicated and justified. A good agency knows that efficiency at the expense of billables is required to maintain competitiveness and good standing with clients, and communicates this to their employees. This model allows for changes in scope with fluidity and active consultation of cost/benefits for a given requirement.
An absolute requirement for this to work is trust between the agency and company. It is incumbent upon the agency to tirelessly work pre and post sale to engender this trust.
This requires candor and constant communication. It also can require occasional empathy and understanding when a mistake is made. That can mean eating additional cost, like too much time used in research or unexpected hiccups due to no fault of the client.
Depending on the structure of the contract, a T&M model may be easier to walk away from than a fixed-bid model that may have big milestone payments or deposits.
Value pricing is an innovative way of pricing that supposedly helps the agency make tons of money, and clients love you, and the tears you weep in joy are licked away by baby unicorns. I’ve read a ton of books on the subject and talked with adherents at conferences, but still can’t square the circle.
The basics make awesome logical sense—agencies should be paid according to the value they provide their clients. If my super knowledge of SEO makes you $2 million of profit with 2 hours of my time, why should I get paid just a few hundred bucks?
The best and most awesomely successful example of this is the $1.5m logo Paula Sheer made for Citibank on a napkin in 5 minutes. In her own words, “It took me a few seconds to draw it, but it took me 34 years to learn how to draw it in a few seconds.”
The principles are simple—if the agency will provide value X to a client, then the agency should be entitled to some portion of that. The devil, of course, is in the details.
How does one accurately and a priori determine the value that will be provided? I’ve never seen any systematic methodology that makes me view the establishment of value as anything more than guessing at what you can convince the client to pay. This can work for hyperspecific niches or when you’re a minor celebrity in the field. You can name your price without regard to hourly rate or margin when a client is approaching you as a top designer. The arguments for value pricing all seem heavily coming from the perspective of the agency/designer, not the client, as they all involve the client paying more for the same work. This works fine when there are no good substitutes, but if there are other, equally talented agencies that do not decide to apply a 80% margin, the client would see better ROI with the agency that doesn’t use value-based pricing.
A retainer is a model that’s more common with old-school and PR agencies (or at least huge and broad retainers). Honestly, as a digital agency mostly using time and materials, it was a few years before we had to seriously consider the logistics of this model. The idea is that you reserve a block of time from an agency in exchange for a modest discount and guarantee of hours/costs on a monthly basis. Typically the hours are use-it-or-lose-it, and the agreement runs for some period of time (i.e. 6+ months).
At S4, we’ve moved into a model that combines retainers with agile project management principles for many of our clients. Agile was born out of software development project management and, while we do use agile for software development, we’ve also begun using it for broader marketing initiatives with clients. Through research, discovery, and past experience, we build a massive backlog of tasks that reflects our strategic recommendations. We then move into two-week sprints, each sprint with a fixed cost. Each task is estimated and prioritized based on the success metric we are focusing on (i.e. organic traffic, conversion, retention, etc). Prioritized tasks are pulled into a sprint until it is ‘full’ based on the budget. Over the following two weeks, we execute on that sprint and build future sprints.
This model allows us to keep working toward measurable and agreed-upon goals while fixing costs to the client and revenue to the agency. The size of the budget dictates ‘velocity’ or how quickly we can execute on the tasks in the backlog.
Performance pricing is an attractive model—you negotiate a base price and then add in incentives to increase profitability as certain targets are hit. We have discussed this with a number of clients, but so far have not pulled the trigger. It’s my opinion that you need to have an existing relationship with the client so you know how they are to work with. In addition, you need to believe in their product, delivery, and organization. At that point, performance pricing should be based on a campaign that you can test the waters with. I love the idea of putting your agency on the line to get results. However, the reality is you need to have a strong two-way trust, as your compensation will rely on numbers provided back to the agency.
This will depend on the specifics of what you’re looking to do and the characteristics of your company. Each model was designed to address a set of problems and carries its own unique risk profile:
FIXED-BID: More certainty about the deliverable, but usually is less able to be easily or cheaply terminated.
TIME AND MATERIALS: More flexible, but there’s less certainty about final cost. Often, it can be terminated at any time given you pay for work that has been done. Requires a higher degree of trust in the agency than some other models.
VALUE PRICING: Bit of a wild card, as I believe there’s some variation on how agencies implement it. Typically, margins are so high the agency has room to ‘eat costs’ and still make a sizable profit. However, I’d pay special attention to the contractual details here.
RETAINER: Somewhat like T&M, you’re just getting a guarantee of effort not of outcomes. Could be difficult or costly to cancel based on the termination provision in the contract.
SPRINT-BASED PRICING: As a variant of the retainer model, the risk profile is similar. At S4, we begin an engagement with a 6-month contract and then go sprint to sprint allowing the client to cancel if they’re not seeing value.
PERFORMANCE PRICING: Carries a low-risk profile, as you’re paying for outcomes not effort. However, can be difficult to structure initially. Requires trust about accurate reporting between the two parties.
Personally, I’m not a fan of the fixed-bid model for anything but the simplest projects. I am also leery of value pricing, as it seems more focused on extracting the largest amount of revenue from clients as a primary principle. At S4, we’ve heavily used time and materials for the majority of our 10-year history. T&M works great but requires trust between both parties. It also works best when there is good communication and project management on the agency side, and when the client makes a point to pay attention to all communication from the agency. We’ve seen some success with retainer/sprint models that addresses issues with T&M, such as reducing variability between billing cycles.
Understanding some of the different types of estimates marketing agencies commonly provide puts you in a stronger position during the sales process. In addition, you will be able to better compare different types of quotes from different agencies.