Has anyone set something like that up? If so, do you consider it worthwhile?
Has anyone set something like that up? If so, do you consider it worthwhile?
We're in the process of building such a solution... lots of work, but we're expecting that the benefit will more than make up for it. For our purposes, external solutions do not support our individual needs - not to mention that they're oftentimes quite expensive.
Why build it yourself?
No single one could do what I wanted. I wish I could do it myself.
I looked into outsourcing a team of programmers but how do you trust them? I am between making something for my business or creating a product that would be fully supported and sold on the market.
You know when you know what you want… and you know it would be a brilliant thing that thousands of people would simply love.
Coding is my problem...
The more tightly you're integrated, the most you need ongoing support (and therefore, potentially more expansive). Google's API sunsets a release every 6 weeks or so, and only supports the last 3 releases. Although, there aren't many major releases each year, most are incremental updates.
An example of tight integration would be inventory management to AdWords keywords. Auto pause/start your keywords based upon items in stock.
More generic uses are bid management systems that use conversion tracking placed on your website. This is fairly easy, but doesn't integrate into any of your backend systems.
If you don't know what you're looking for, then 'scope creep' usually becomes an issue along with price. If you know, then the proposals are much more manageable.
How did it go poster_boy? Anybody else try this?
It's set to launch within days... quite exciting. Important to note: It's a substantial amount of work - and requires a key investments of resources, time and patience. Not just the development time... but also the detailed requirements definition - as eWhisper noted, as well...
But, at the end, an internal system that uniquely achieves your goals... in our case, automating rules-based keyword creation and submission (across multiple languages & brands), per keyword inventory management, reporting, optimization rules & alerts, etc. ...is priceless.
I like that approach. Saves a lot of time (later on) if you integrate website-design, warehouse-management, <insert dozens of interesting other intefaces here>, and then adwords into a tight and effective structure, designed for your very own specific needs. click-path-effectiveness. But programming also takes a lot of time.
I had the same idea, but I am relatively new to adwords and for the time being prefer to make the classic beginner-mistakes with the tools available. For instance, I read in a book on adwords that one of the classics is to try to import huge bulks of keyword selections. Instead, some careful finetuning by hand was often the better way, the author says. I must admit, there is some truth in this.
Nevertheless, I believe that long-term it is necessary to built some tools, which will provide some advantages against competitors.
Hope I'll manage to double our organic orders with the help of adwords until the end of the year. I need this target as a proof that adwords in itself is a hopeful path for my small niche in the future.
I had some eperiences with the soap-interface on another project, so the whole thing shouldn't be too complicated; I think in a few months (maybe Feb 09) it is quite likely I need some better tools to manage my account. (Lucky me, I'll only have to manage one.)
If you're really going to begin this in the near future, I'd be very much interested in your experiences with how you designed your database and a brief sketch of some of the things you found worth programming yourself.
I'd say the first thing to do is to set up a database with three tables:
campaigns > adgroups > keywords
just as the adwords structure in itself suggests. Maybe a fourth level, if you manage different accounts for other clients. To sort out such doublettes should be quite easy simply by defining a unique table-row for the keywords you use.
I think the api is best used in order to synchronize this database with your adwords account. to and fro. I'd start with setting up these three tables and then a routine to import adwords-csv-export-file. And then step by step use the api to send modifications (made locally) to the account. But there is a lot of work involved to even catch up with the options the adwords-editor gives. Once you reach that level, true advantages may start...
It's easier to spec out how to bring in data from MSN, Google, and Yahoo and store it before you build a system then to try to hack it together at a later date.
If you manage a lot of keywords and ad copy, I would suggest your warehouse store all your data so you know what should be going on at each vendor, and then sync that data on a regular basis. It's much cheaper (and faster) on the API calls to queue your warehouse for changes rather than to query the vendors.
[edited by: tedster at 11:58 am (utc) on Sep. 9, 2008]
