this post was submitted on 20 Oct 2024
281 points (95.5% liked)

Ask Lemmy

26916 readers
1582 users here now

A Fediverse community for open-ended, thought provoking questions

Please don't post about US Politics. If you need to do this, try !politicaldiscussion@lemmy.world


Rules: (interactive)


1) Be nice and; have funDoxxing, trolling, sealioning, racism, and toxicity are not welcomed in AskLemmy. Remember what your mother said: if you can't say something nice, don't say anything at all. In addition, the site-wide Lemmy.world terms of service also apply here. Please familiarize yourself with them


2) All posts must end with a '?'This is sort of like Jeopardy. Please phrase all post titles in the form of a proper question ending with ?


3) No spamPlease do not flood the community with nonsense. Actual suspected spammers will be banned on site. No astroturfing.


4) NSFW is okay, within reasonJust remember to tag posts with either a content warning or a [NSFW] tag. Overtly sexual posts are not allowed, please direct them to either !asklemmyafterdark@lemmy.world or !asklemmynsfw@lemmynsfw.com. NSFW comments should be restricted to posts tagged [NSFW].


5) This is not a support community.
It is not a place for 'how do I?', type questions. If you have any questions regarding the site itself or would like to report a community, please direct them to Lemmy.world Support or email info@lemmy.world. For other questions check our partnered communities list, or use the search function.


Reminder: The terms of service apply here too.

Partnered Communities:

Tech Support

No Stupid Questions

You Should Know

Reddit

Jokes

Ask Ouija


Logo design credit goes to: tubbadu


founded 1 year ago
MODERATORS
 

I work at a place where data quality is not on anyone's radar. We have a reporting team in our group so we do our best where we can, but combining any datasets with other groups (like marketing & sales) is next to impossible as each team is silo'd and do things their own way - think free-form text fields to tag content...

How can I politely and succinctly say the above? Also, anyone else in a similar boat?

top 32 comments
sorted by: hot top controversial new old
[–] partial_accumen@lemmy.world 167 points 4 weeks ago* (last edited 4 weeks ago) (3 children)

Implementation of AI requires strong unified data governance and data hygiene to produce company wide strategic solutions. The current company posture instead is focused on tactical level data collection and analysis which does not lend itself to consumption in relevant possible cross-department opportunities.

[–] SzethFriendOfNimi@lemmy.world 74 points 4 weeks ago* (last edited 4 weeks ago)

And if you want to tank it without overtly tanking it.

“We will need to establish a review and governance board to establish standard data structures and reporting that can be used to drive the initiative.

It will need to be cross team and cross specialty so we should start by establishing a group to identify those people so we can proceed”

A year later and you’ll be lucky if they’ve even picked out who can be part of the review process let alone agree on some convention and adjusting their tooling and processes to make that work.

[–] Noxious@fedia.io 19 points 4 weeks ago

This is gold

[–] NocturnalMorning@lemmy.world 12 points 4 weeks ago (1 children)

Somebody give this guy a promotion.

[–] PlasticExistence@lemmy.world 14 points 4 weeks ago (1 children)

I dunno. They forgot the hyphen between company and wide. 3/10.

[–] brbposting@sh.itjust.works 10 points 4 weeks ago

Not three out of ten: forgetting a hyphen is unforgivable & puts it into negative territory. I would drop it by six points from there and give it 3/10.

[–] Fontasia@feddit.nl 40 points 4 weeks ago

They're not wanting to work better, they just want to be able to say they "use AI".

Get them an estimate of how much it would be to give everyone Copilot for Business licenses and they will start talking about the business having other priorities.

[–] akilou@sh.itjust.works 40 points 4 weeks ago (1 children)

We're unable to leverage some of the latest advances in AI due to our Leadership's abundance of caution and strategically allowable risk profile.

[–] hddsx@lemmy.ca 9 points 4 weeks ago

You can effectively shorten the second part to “due to the company’s allowable risk profile”

[–] PriorityMotif@lemmy.world 30 points 4 weeks ago* (last edited 4 weeks ago)

"Each department in this company has developed their own way of doing things. Unfortunately, not every department is giving us information in a way that we can use it to our advantage. We need to get everyone across the company on the same standards so we don't have any mixups. We can't help you do your jobs better if we can't process the information you give us."

[–] Greg@lemmy.ca 29 points 4 weeks ago (1 children)

We need shared definitions to tell meaningful stories with our data. And then use a company specific example like how a customer's journey can not be understood with differing definition between marketing and sales. The marketing team can't measure the quality of the leads they're producing unless they can directly link a customer's whole journey from acquisition to churn. Otherwise it's just vanity metrics. But don't be too harsh, vanity metrics are really common in business. A company needs strong data leadership to create a culture of using data to justify decisions to a culture of using data to inform decisions.

[–] NOT_RICK@lemmy.world 11 points 4 weeks ago* (last edited 4 weeks ago)

Definitely try to use examples to help them get a glimpse into the issue. I like to explain documentation errors by pointing out when what are supposed to be sequentially recorded timestamps are recorded out of order in my work’s database. Sometimes the data quality isn’t there.

[–] Grimy@lemmy.world 24 points 4 weeks ago* (last edited 4 weeks ago)

Our team does its best to maintain consistency within our own processes, but collaborating with other groups like marketing and sales can be challenging due to siloed operations. Each team follows different approaches to data management, often using free-form fields and inconsistent tagging methods, making it difficult to combine datasets effectively.

You pretty much already had it.

[–] GBU_28@lemm.ee 22 points 4 weeks ago (1 children)

"sure thing boss"

Starts looking for new job

[–] OpenStars@discuss.online 5 points 4 weeks ago (1 children)
[–] GBU_28@lemm.ee 5 points 4 weeks ago
[–] Etterra@lemmy.world 21 points 4 weeks ago

Having been in a situation in which management could not quite grasp the concept of "what you're asking is literally impossible" the only answer I have to give is to leave and find someplace smarter to work.

[–] Konstant@lemmy.world 13 points 3 weeks ago

This is what AI is for. Why don't you ask ChatGPT?

[–] Boozilla@lemmy.world 13 points 4 weeks ago

Resign, then offer them consulting on it for $300/hour.

[–] wjrii@lemmy.world 12 points 4 weeks ago

Our legacy systems have served us well. Unfortunately, the facts of being market leaders and first movers mean we're not operationally equipped to leverage turnkey AI solutions. We will need significant buy-in from management in several stakeholder organizations, as well as significant time and resources for procurement (or development), implementation, and change management.

(optional depending on your GAF level) : With the right level-set and commitment, integrating our systems for AI could be transformative in the best ways.

[–] sunbrrnslapper@lemmy.world 10 points 4 weeks ago

It is actually easier to just get a different job with better leadership. You can say exactly what you wrote in your exit interview.

[–] monsterpiece42@reddthat.com 10 points 3 weeks ago (1 children)

www.goblin.tools is a godsend for this type of stuff.

Here's what it put out using the "Formalizer" tool set to "More professional":

Our leadership team primarily consists of individuals with extensive experience, but there appears to be a gap in understanding contemporary data practices. This challenge makes it difficult to fully leverage the potential of AI in our operations.

[–] ramsgrl909@lemmy.world 2 points 3 weeks ago (1 children)

Have been using this the last 2 days and love it, thank you!

[–] monsterpiece42@reddthat.com 3 points 3 weeks ago

Hell yeah! Happy to help--it's definitely a game changer for me and plenty of others. I wish you the best luck moving forward!

PS there is an app, and I think it's $1. Worth throwing the dev something for their efforts imo!

[–] blackbrook@mander.xyz 9 points 4 weeks ago* (last edited 4 weeks ago)

Removing the unecessary adjective "old" would be an obvious start.

But what is your audience and what do you hope to accomplish? Unless you really have an audience who is likely to accept that the people in leadership are lacking an important competence and the power to go around them to get something done, then I would assume that leadership is your audience. In which case the part that blames them is entirely unhelpful and you should stick to explaining the problem and needed changes.

[–] daddy32@lemmy.world 9 points 3 weeks ago

Well there's a classic saying for this purpose: "Garbage in, garbage out".

[–] queermunist@lemmy.ml 9 points 4 weeks ago

Why would you even bother?

“The org is extremely top-heavy”

[–] IMALlama@lemmy.world 5 points 4 weeks ago

Good luck. The company I work at has the exact same problem. Since each system tends to be owned by a different org, and the systems all meet the owning org's needs, you're going to be in for struggle.

[–] souperk@reddthat.com 4 points 3 weeks ago (1 children)

I know you are asking for something different, but since there are already a few good answers, allow me to instead to reject the premise and give you a different.

It's not impossible to implement an AI solution within the context your provided. The problem is that it's going to be expensive. However, you can offer to deliver something smaller, focus on the smallest but valuable contribution you can make. While cleaning up the data is still going to be a hell of task, if the scope is small enough it can be achievable. Then, you can communicate the difficulty to scale due to data issues which can help management undestand the importance of prioritizing data quality.

If you have a bunch of sales data, maybe you can focus on deriving purchase patterns and build a simple recommendations engine. If you want to focus on marketing, you could try lead classification. Ideas depend on the domain of the company you work for.

[–] dumples@midwest.social 6 points 3 weeks ago

If you have a bunch of sales data, maybe you can focus on deriving purchase patterns and build a simple recommendations engine. If you want to focus on marketing, you could try lead classification. Ideas depend on the domain of the company you work for

This is where we get the fun part of definitions. Depending on what people think AI is this aren't AI. Most people mean GEN-AI aka the new fancy shiny thing. These are boring old machine learning, data science, statistical learning, data mining etc. (depending on your definition)

[–] HobbitFoot@thelemmy.club 3 points 3 weeks ago

AI can probably help with formatting issues, but what are you intending to use AI for?

Most of the companies that have used AI successfully have been ones with targeted goals for their AI. That your company has siloed information buckets means that each division seems to generally be operating by themselves. What is going to happen when a computer output says you need to tweak how the company is running?