Let's chat! When’s a good time?
Thank you! Your submission has been received!
Oops! Something went wrong while submitting the form.
The following is the second half of a two-part series about the factors involved in building and buying a B2B marketing analytics and attribution solution. This post deals with the cost and time requirements for an in-house and an off-the-shelf solution. It also compares the opportunity costs of building and buying a solution.
Be sure to check out part one which talks about the need for a B2B marketing analytics and revenue attribution solution. Along with a breakdown of the technical requirements for each solution.
This segment is an overview of the costs involved in building an in-house B2B marketing analytics and revenue attribution solution. While there are certainly other costs involves, we cover 5 of the most prominent ones:
· Cost of ETL
· Cost of Data Warehousing
· Cost of Data Processing
· Cost of Data Visualization
· Cost of Staff
Extract, Transform and Load (ETL). Extracting structured or unstructured data from a source — this could be data from your CRM or Google Ads. Transforming includes processes like cleaning, duplication, sorting, etc. and ensuring data integrity and compatibility. Loading involves placing all of the transformed data into a data repository or a data warehouse. The data could be either loaded completely or at predetermined intervals.
This is the cost of maintaining a data pipeline. While it is possible for your engineering team to set up a data pipeline, some companies find it cost effective to use an ETL tool. These tools include tools such as Hevo, Fivetran, Google Dataflow, Pentaho, etc.
Fivetran, foe example, lists a range of pricing tiers. The starter tier’s price (which is for a small team’s data stack or the bare minimum for an in-house solution), would depend on the number of rows you update. This will range anywhere from 2 million total rows at a monthly estimated fee of $120, to 500 million total rows at an estimated $4,628 per month.
In the previous blog we talked about the need for a data warehouse from an analytics perspective. We discussed why relying on application databases is not scalable. While you could invest in a local data warehouse, there are a multitude of benefits to investing in a cloud data warehouse. Ultimately, this will prove to be convenient when building an in-house solution. Scaling operations would be expensive as it requires more ram (and in all likelihood, a dedicated database manager). That being said, cloud-based data warehouses like Google Cloud Storage, AWS Redshift, Microsoft Azure, and Snowflake fit the bill.
Cloud data warehouse storage prices vary depending on a range of factors. Google Cloud Storage, for example, has options varying in region and class — the class of storage, standard, nearline, coldline, and archive is determined by the frequency of access to the storage. In the region of US-central Iowa, at the standard class, warehousing will run you about $0.020 per GB per month.
Most cloud-based data warehouse services also include processing data. This is the cost to process SQL queries, scripts, functions, and more. This is in addition to the cost of loading data that you are processing in storage. Processing data is usually handled by a database management system like Bigquery, AWS Redshift, Oracle, Singlestore, etc. These services offer Cloud database as a service.
The cost involved in the pricing of these services includes the use of vCPU, Memory and cloud storage. Singlestore, for example, on its standard plan has a starting price of $0.65 per hour and will increase depending on the number of vCPUs and memory used. A vCPU of 16 and 128GB of memory will cost you $3,796 per month.
In the previous blog, we talked about the presentation of your reports to your end-user. This requires a data visualization tool. A skilled engineer could purchase data visualization libraries and build them out. But for the sake of time, a lot of businesses resort to data visualization tools like Tableau, Looker, and PowerBI.
A data visualization tool like Tableau will cost you $70 per month per license.
Staff will, by far, be your most expensive costs. To build a marketing analytics and attribution in-house solution, you would at the very least require a small team of 3 full time data engineers and 1 data scientist. You will require experts with experience across programming language and ETL. In the US, the break down is as follows: on average a data engineer’s CTC is $116,772 per annum, along with a $5,000 cash bonus and other non-cash benefits as of 2022. The average CTC of a data scientist in the US will cost $102,865 as of 2022. (Indeed.com). These costs will have to be multiplied by the number of data engineers and scientists hired.
In terms of cost an off-the-shelf solution like Factor.ai will as of this date cost you $1,188 per annum on the starter plan which includes web analytics, multi-touch attribution, funnel mapping, Metric reports and more. Their growth plan on the other hand will cost you $5,988 an AI powered “Explain feature”, automated weekly insights and a dedicated customer success manager
To build a fully operational in-house B2B marketing analytics and revenue attribution solution, with a team of 3 full time data engineers and 1 data scientist will take anywhere between 9 to 12 months.
An off-the-shelf solution like Factors.ai can be set-up in minutes. It requires no professional services for onboarding either.
Now, we're all caught up about the resources required to build an in-house B2B marketing analytics and attribution solution, as well as what to expect from an off-the-shelf solution. So should you build or buy? This section runs through the opportunity cost of building and buying. Essentially, what are you missing out by choosing whether to build or buy.
By choosing to build an in-house solution you forgo the benefit of:
· The cost savings earned from buying a solution
· The time saved from not having to set up an in-house solution
· No-code integrations and developer dependency
· Maintenance and innovations handled by the service
· Using an advanced SDK, and not having to optimize SDK
· Data cleansing handled by the service
· Data visualization within the same product
· Unified Dashboard
By purchasing an off-the-shelf solution you incur the following opportunity costs:
· Product may not fulfill very unique analytics aspects of your business
· Product may not deliver on their promises
· Certain products may not fulfill your data privacy requirements (learn more)
· If a vendor liquidates or gets acquired, you cannot ensure data ownership and continuity of business
The most important point to take away from this is that when you build an in-house solution, you would have to weigh the risk of doing so. The average tenure of a CMO is about 40 months. Would they prefer to spend the first 9 to 12 months of their tenure waiting on a solution that isn’t proven to meet their need, or have a solution that is up and running within a week for a fraction of the cost of building one?
In my opinion there is too little to gain and a lot to lose when buying. Most of the opportunity costs of buying could be avoided with modern solutions like Factors.ai. Where custom plans can be built to fulfill your business’s unique needs. A demo of the product can be requested to ensure if the product delivers on its promise. Factors.ai uses first party cookies and is GDPR, CCPA, PECR and SOC2 compliant. And Factors.ai can send their client’s data to their Bigquery instance on demand giving full data ownership to the client.
Still on the fence? Book a demo with Factors.ai now.
Get the latest best practices in Marketing Analytics
delivered to your inbox. You don't want to miss this!!
.webp)
