4 ms·
I perform data cleansing of excel spreadsheets and sometimes convince them to develop basic CRUD apps to take the place of excel. Depending on the project and t
by mro_master 7y ago
I perform data cleansing of excel spreadsheets and sometimes convince them to develop basic CRUD apps to take the place of excel. Depending on the project and the feeling I get about the depth of their budget I charge: per row based on the transformations they need, flat fee if they actually have well defined requirements, per hour for development time if they are convinced of the value in standing up a new web app to handle the process.
- opportune 7y agoHow do you get clients for this kind of work and about how much do you charge? This sounds pretty laid back and potentially lucrative
- mro_master 7y agoThat my friend is the secret sauce. My first job out of school was for a staffing company that pimped me out to perform similar projects. Word of mouth mostly, it's very lucrative if you have domain knowledge and can sit for long periods of time. It's not Mechanical Turk style cleansing and I have to take personal responsibility for every row as they mostly just suck up my cleansed data and roll with it, which you can't do with ML cleansed data or off-shored cleansed data. I've been thinking of launching a SaaS product but the transition from trusted consultant to enterprise SaaS is not easy. What do you do?
- opportune 7y agoI do big data stuff mostly. A lot of that involves cleansing. So working with large excel/csv/other tabular data formats is second nature to me. Familiar with many technical tools that I could maybe use as force-multiplier (e.g. serverless SQL processing, OLAP, RDBMS, python pandas, spark), hence my interest since if we relax the condition that every single row is accounted for, my productivity would likely be pretty high
- mro_master 7y agoThe customers always find the rows that are not accounted for, it's bad form to drop data. We sell high quality data, which you can't claim if you can't account for every single row. You should be careful about delivering an incomplete data-set.
- shapiro92 7y agoi did the same but added to the mix data acquisition aka crawling or buying from someone else and re-selling. It can be tedious sometimes but boy it did pay well. I stopped as I decided to get back into the workforce instead of going solo.