As a sibling, I have had a lifetime’s worth of experience with attribution. Broken lamp? That’s attributed to my brother. Leftover candy after Halloween? I’m the first one to see it sitting by the door, so I get the Skittles conversion. 
In marketing, attribution modeling is determining where a conversion action came from. Whether it is an ecommerce sale, lead acquisition, or other interaction with your website, it is very important to know which marketing channels are responsible for your goal completion. Knowing where your revenue is coming from will let you maximize your investment in each channel and increase the subsequent return.
Attribution modeling in Google Analytics will be different based on your company’s specific goals. To help with determining an effective attribution model and how to most effectively use the attribution model, here are some current options that you can use in GA:
Last Interaction - All conversions attributed to the last channel used. A company that only uses paid channels like AdWords or offline media buys might want to know which marketing cost is bringing the most return on investment.
First Interaction - All conversions attributed to the first interaction that a user has with your business. This is a good option for lead management if your sales team has a different approach based on how the user initially interacts. A lead calling in for general information might be nurtured in a different way than a user submitting a quote request for materials.
Last Non-Direct Click - Ignoring an interaction being a direct visit to your site, all conversions are attributed to the channel used before the direct visit occurred. If your business is very well known, you might want to ignore visits as direct and instead focus on the other channels they used that ended with that direct visit.
Last AdWords Click - All conversions are attributed to the most recent AdWords click. I like this one and I am in no way partial as a representative of Paid Search. Since this will attribute more conversions to AdWords that might have actually converted on an email touch or direct visit, a company would use this attribution if a vast majority of effort and spend involved Paid Search.
Linear - Each channel in linear attribution gets an equal piece of the conversion distribution. A company using a cross-channel, unified marketing strategy would benefit from linear attribution as the conversion goal would be the same regardless of how the user interacted.
Time Decay - Uses a half-life of a week to give more credit to a channel if the interaction is immediate and less credit as time passes. A company sending a blast email might consider using a time decay model if the email campaign is time sensitive or if a conversion is worth more when it occurs sooner rather than later.
Position Based - A custom hybrid of Last and First, allowing a percentage of attribution to be assigned to first Interaction, Last Interaction, and middle interactions. A company can set a higher percentage of the attribution to a first click interaction and a last click interaction if they value how a prospect initially discovers the company but are still interested in showing how the prospect converts.
Each of these attribution models will offer a different view, so it will be important to keep your own goals in mind when you determine how to attribute an action to a marketing channel.
This is not the greatest blog post in the world, no, this is just attribute.