The Case of Gemini for Google Workspace
The Hard Sell
Today I received an email from Google Workspace, urging me, in my capacity as admin for TheRoadTLV.com, to activate a "14-day Gemini for Workspace trial".
The Debacle
I'm old enough, though, to remember the Gemini ai debacle, back in (checking my notes) February this year, AD2024.
Not that trust instilling, was it? Do I want to invest time into this, risking my team integrity and reputation? Nah...
The way I read it is:
"Do you remember that snafu we did 9 months ago? Why don't YOU try it?"
Closing this email, never to even think about it again (well, except for this post).
Is it all lost, then?
This email is a classic SLG (Sales Led Growth) move within a 'Land and Expand' strategy. Since I am an existing (purportedly happy) customer, there should be enough trust for me to give a look at advances of this type.
The thing is, Google did blew this, as it comes to Trust. Not a shallow pit to emerge from...
Is it all lost then?
Two wrongs don't make a right. Unless...
This got me thinking. I just recalled a ubiquitous feature in the Workspace I was less impressed from - Help.
Isn't the help here so 1999? I mean, it looks like a google search over the knowledge base, without much relevance scoring. When I follow the links, nothing much comes of it, I'm afraid. Unless I hit an exact search terms, I'm bound to see results such as this one:
You'd be the judge of the quality of this result in terms of relevance and accuracy. I was less impressed.
Apparently I'm not the only one underwhelmed. ChatGPT finds on 6-7 score (out of 10) user satisfaction for this feature, with
Limited personalization
Difficulty finding detailed, advanced solutions
Over-reliance on community answers
as the main negative factors.
A more source traceable Perplexity cites the following negative factors:
Complexity for new users
Limited self-service options
(and ironically) Search functionality within the Help Center could be improved
So we have two wrongs: Poorly implemented help feature, worthy of an earlier era, and a magestically failed new ai service, that badly needs credibility - both on user experience and result quality.
Enters PLG to Regain Trust
Mmm... isn't it a tough one? what if we had a large language model we could train on the queries users ask, the knowledge base, and perhaps support sessions?
You say we do? "Gemini" you say? Interesting...
Here's a PLG (Product Led Growth) strategy that might save the day, I believe, in that it helps regain trust.
It's simple:
Train Gemini on all things help and support across Google Workspace
Avail it to users gratis to let them experience the quality (beware, though, can't botch this one!)
Offer whatever you want from within this flow
Voilà!
My assumption is that administrators are both decision makers and heavy users of the help feature. Google can now prove users Gemini's worth, (re)gain their trust, and negotiate its way as a viable ai tool for teams using the Work.
What's Next?
The next steps would be to build and test this help-chat functionality. KPI's to look for:
Dropdown Open: When the Drop down menu is pulled open
Dropdown Select: When the chat window is launched
Text Input: When users type in the start of the conversation
...And so on, depending on the flows that still need to be developed...
What's your take on this?
Always happy to discuss, share ideas. Drop me a word at yoel@theroadtlv.com!
Komentar