Analytics Superpowers - KPI Analysis

[@DaveOnData]

👩‍🔬👨‍🔬 Learn to use your basic Excel skills to analyze the business like a Facebook data scientist: 📺 kzbin.info/www/bejne/rnq7qqCXl955f68

[@PillsLifestyleReviews]

Thank god I found your channel Dave. I've been learning data analysis at variable pace for about a year. I've been learning the tools, the math, and how to clean data. Now I'm starting to get a feel of how to actually analyze data. I couldn't find what I was looking for on the more famous KZbin channels. Your pretty obscure channel on the other hand has the legit info!

[@debjaniguha1345]

KPI Analysis is indeed the most important thing in the entire field of Data Analytics. In the sea of tools like Tableau, Power BI, excel etc. we forget the basic objective why all these tools were developed and what is the objective that we are trying to achieve. It is KPI analysis and identifying a problem area that is core to any data driven decision.

[@DaveOnData]

Thank you for taking the time to comment and add your perspective to the conversation! So glad to hear that I'm not the only one that thinks KPI analysis is integral to data driven decision making.

[@RichardDurham]

Process Behaviour Chart for the win! That said there is more to using PBCs so that you stop reacting to usual variations yet highlight variations that signal changes to the process that require investigation.

[@DaveOnData]

Absolutely! That's one of the major benefit of PBCs - the detections rules that standardize separating the business signals from the noise.

[@ahamedazam3168]

Hi David, Just want to say that you are truly amazing! So many practical things to learn based on real world cases! Really appreciate each of your videos.

[@VhangorD2]

great video! KPI analysis is exactly what I need to starting to merge my knewly acquired aknowledge in data analytics and my current job.

[@DaveOnData]

Excellent! So glad you found the content useful.

[@jp410]

What I like about your channel its practical orientation, and the implicit skepticism towards business lingo and over-complication. I did small-scale consulting and programming work 20+ years ago -- VB, SQL, etc. I did requirements gathering (interviewing users, managers), then prototyping, then DB and front-end programming and implementation. Then I did something really stupid -- I went and got a PhD in Cultural Anthropology from Berkeley. I never got an academic job, and found that I had basically educated myself out of the industry job market, as my skills had become outdated, and academics, the story goes, are neither practical nor deadline-driven. I'm horrible at job-seeking, and I get put off by snake-oil lingo that business cultivates. Your videos give me the confidence to begin to connect the latent skills I have (which don't seem that advanced to me) to the current state of business IT (which always demands that things sound advanced).

[@shisk1]

Awesome video! Thanks for sharing your knowledge David!

[@DaveOnData]

Thank you so much for the feedback and glad that you found the video useful!

[@farahdk1398]

Thank you for this valuable info on the process behavior charts!

[@DaveOnData]

You are welcome! Glad you liked the video.

[@icybrain8943]

Reminds me of how we monitor analytical equipment in the medical laboratory using Westgard rules.

[@DaveOnData]

Yes! Process behavior charts is another name for XmR charts from the area of statistical process control.

[@datascienceharp]

Amazing resource, thanks Dave!

[@DaveOnData]

Thank you, the feedback is much appreciated.

[@ihabmusslih7520]

you are fantastic , good job

[@uwillnevahno6837]

Couldn't the guardrails 11:08 be built into the dashboard's upper right visualization 4:18? Something along the lines of "If total 50 then change color to red" to alert the manager to investigate why there are too few or too many staff. There might also be some seasonality to the peaks and troughs.

[@DaveOnData]

If I understand your question correctly, you can automate the color-coding of the data points based on what are known as "detection rules." However, there are 3 detection rules, and this is the simplest of the three to implement (i.e., a data point in the "danger zone"). Seasonality is also a big concern, as you correctly pointed out. In the case of a call center, it might require looking at call volumes at the weekly grain to avoid within-week seasonality (e.g., calls decrease each weekend).

[@bernardogarcia8607]

Learn something new. Thanks

[@hamed19752]

A lot of repetitive information, possibly could have been explained in 3-4 minutes.

[@kwamelegend3587]

I have a question. Do you automate your reports ? When you create a dashboard for a manager, do you automate the process so it automatically updates ? If so, how do you automate the process? Do you use excel?

[@jourdango2615]

Power Query or Power BI is what you need

[@ShivaniDavne]

wasn't really very helpful. It would be great if you were more precise. There is a lot of irrelevant story telling rather than focusing on the particular point. I am saying this only after watching 4-5 video of yours. Just trying to make it better, please take it as a feedback!

[@DaveOnData]

Thank you for feedback! Hopefully you will find my other videos more useful.