Exploring GPT Productivity Hacks
Being a business owner, especially mainstreet business owners can mean having a lot to do with so little time available. Even with all the applications and tools currently familiar to us, we still feel like we’re racing against time to get everything done. Let’s face it, aside from the important things we have already budgeted…
Being a business owner, especially mainstreet business owners can mean having a lot to do with so little time available. Even with all the applications and tools currently familiar to us, we still feel like we’re racing against time to get everything done. Let’s face it, aside from the important things we have already budgeted for outsourcing, there are still aspects in our business which we wish could be delegated to someone else. Today I’m going to show you 1 (and a half) tool that can be helpful for improving your business productivity!
Introducing: GPT Tool
OpenAI is an AI-research company that provides AI tools that are user-friendly to everyday users. Part of their research and development was on creating GPT: a tool based on an AI model designed to help with improving and assessing language. That means we now have a tool that can work on understanding human language and its nuances and creating effective responses from the data it has collected. Simply put, communication is largely a part of what makes or breaks a business. Every aspect of business involves a connection between two individuals and how one effectively communicates determines the success or failure of a transaction. In fact marketing is largely founded on good communication. So if you can somehow find a way to use the GPT tool to improve your communication processes in the company, you can imagine how much of a transformation your company can undergo with assistance from artificial intelligence.
There are two models you can use for free right now:
- GPT playground (based on GPT3)
- More Sophisticated language model that can really do amazing things
- Really popular now, and is essentially built on top of the playground model
- Given a chat interface so it’s more like engaging with an assistant
Playing Around the Playground vs ChatGPT
If you were to give a request both to Playground and ChatGPT, it is expected that there will be similarity to both models however you will find these key distinctions:
|More thorough||More simplified for conversation format|
|More stable||Can crash sometimes|
Where does this all come into importance when applying in your daily use? Say for example there’s a rather interesting article that is around 1000 words or even more. You know you will have to spend a good 30 minutes or even more to really absorb and pick up the most important highlights of the article. You might even want to assess first if this article is worth your reading time because, yes, time is our very limited resource. Simply request from GPT a summary of the article and from there you can have a general idea of what to expect from the article. No need to skim through the words and jump around paragraphs, the AI model will do that for you. Sort of like when you were asked to write a book report for a certain novel back in school but you tried to find some cliff notes or summary notes that will help you get through the important details of the book assigned to you.
While that is helpful in a very basic task, you can also see how it can help with certain tasks involved in digital marketing such as content writing. In SEO (search engine optimization) you are asked to provide a meta description that helps Google and other search engines understand what your content is all about so they can index you better in their search results. Instead of spending time trying to write the best description that you think an algorithm can process, why not let an artificial intelligence write something another artificial intelligence will process effectively.
Simply type in the GPT playground: “Please write me a good meta description that is no longer than 200 characters and will encourage the reader to click read more. Here is the text: (insert text here)”
Take note, you don’t even have to copy paste entire texts. You can simply feed it URLs of blog posts or websites and they can extract the data themselves to process your request.
Now we know that we can produce quality micro-content easily from the GPT tool, we can begin to explore where in our operations we find ourselves feeling like we don’t have enough time like creating social media posts.
For example, if your business has great potential to do brand presence and lead-generation in twitter, you can simply input a link to a certain blog you wrote and request: “Can you write me 3 tweets based on this article from (insert link)”
In a span of just a few seconds, we are able to create social media content without even spending the time to figure out what words to use and how to write the content to make it easily comprehensible for our audience! Of course there are social media marketing best practices you do need to consider but the use of GPT in the long run for these strategies will evolve. As of now, keeping it simple and working with what can be used at the moment is the best way to utilize the opportunities from GPT.
How Much Time Are Spending on Emails?
Funny how emails were supposed to be created to optimize communication so business transactions can be done faster. Today, we know millions out there spend several hours a day going through email after email, constructing them in a careful manner to ensure you get the message across correctly.
Now because emails have a conversational nature, this is where ChatGPT might play a better role in producing responses. Here is an example of request you can type in ChatGPT to produce for you an email to re-engage with a customer you talked to last quarter:
“Please help me write an email to follow up a prospect about my Catering services. We haven’t talked in a few months about this and I want to bring it back up and see if it’s something they want to explore.”
Following the prompt above, this is what ChatGPT provided for me:
Subject: Following Up on Our Catering Services Discussion
Dear [Prospect Name],
I hope this email finds you well. It’s been a few months since we last spoke about our catering services, and I wanted to reach out and see if you are still interested in exploring what we have to offer.
At [Your Company Name], we specialize in providing delicious, high-quality catering services for events of all sizes. Whether you are planning a corporate event, a wedding, or a private party, our team can work with you to create the perfect menu and ensure that your guests are well-fed and satisfied.
I understand that you may have a busy schedule, but I believe that our catering services could be a valuable asset to your next event. If you have a moment, I would love to set up a time to discuss this further and answer any questions you may have.
Thank you for considering [Your Company Name]. I look forward to hearing back from you soon.
Best regards, [Your Name]
Now let’s compare the output from GPT playground which has a more direct approach and a lesser consideration for the conversational nature of the response:
Subject: Following Up on Catering Services
I hope this message finds you well. I wanted to follow up with you regarding our conversation several months ago about my catering services.
I’m sure you’re busy, but I wanted to check in and see if you’re still considering having me provide catering services for any upcoming events you may have. I know the past few months have been an adjustment for everyone as we navigate our new normal, so I wanted to make sure you had the opportunity to explore all of your options.
If you’re still interested, I’d love to connect and discuss further. I would also be happy to provide any additional information you may need.
I look forward to hearing from you.
Both results are quite similar but you can see the tonality of the first model: ChatGPT is more preferable than the direct approach of the Playground model.
More than just writing emails for you, it’s also finding the best way to communicate certain things that might feel awkward or sensitive such as price increases in your services, unexpected delays, or even follow-ups in their overdue payments.
And because emails are basically communication tools, you can see how other language AI models can help you in other aspects of communication that are non-digital. For example you can write a prompt:
“How do I tell my employee constructively and nicely that his performance has not improved over the past quarter and he needs to get better or we might terminate his contract. I also want him to feel hopeful after our conversation.”
And with that request, ChatGPT did not disappoint with this thoughtful guide on what to say to the employee:
Dear [Employee Name],
I would like to schedule a meeting with you to discuss your performance over the past quarter. As you know, it is important for us to continually assess and evaluate our employees to ensure that we are meeting our business goals and providing high-quality services to our clients.
I regret to say that I have noticed that your performance has not improved over the past quarter and I would like to have a conversation with you to discuss how we can work together to improve your performance.
It is important to note that while we value your contributions to the team, if your performance does not improve, we may need to consider termination of your contract. However, I also want to emphasize that I am hopeful that with the right support and guidance, you will be able to meet our expectations and grow in your role.
I believe in your potential and I am confident that with hard work and dedication, you can improve your performance and become an even stronger contributor to our team.
Please let me know a time that works for you to schedule our meeting. I look forward to speaking with you soon.
Now if you want this for your personal life, you can also make prompts on how to write an apology to your wife/husband for forgetting your anniversary gift. Essentially, you will no longer be at a loss for words when AI can help you choose the right ones to say.
Extending To Excel
The language model does not simply limit to words alone. You will find it effective when dealing with other applications such as Sheets or Excel. You can write a prompt like:
“I have a google sheet with a table of information, column a is the service name, and the other column which is column b is the price. On a different sheet I need to look up the price of column b based on the service name in column a. How do I do that in Google sheets?”
Using GPT playground gives a very direct answer:
You can use the VLOOKUP function to look up the price from column B based on the service name in column A. The VLOOKUP function syntax is: =VLOOKUP(search-criteria, range, column-index, [is-sorted]). For example, the formula for looking up the price for the service in cell A2 would be: =VLOOKUP(A2, Sheet1!A:B, 2, FALSE). Replace Sheet1 with the name of the sheet that contains the table of information.
Now you have ChatGPT’s response which not only provides an answer, but also provides an example of how to do this. Very interestingly, trying to follow the way we humans teach certain concepts by extending the context of ideas beyond just simple if this then that answers.
To look up the price of column B based on the service name in column A in Google Sheets, you can use the “VLOOKUP” formula. The formula syntax is as follows:
“lookup_value” refers to the service name you want to find the price for, “table_array” refers to the range of cells that includes both the service names in column A and the prices in column B, “col_index_num” is the column number that contains the price (in this case, it is 2 because column B is the second column), and “[range_lookup]” is an optional argument that specifies whether you want an exact or an approximate match. If you set this argument to TRUE, VLOOKUP will return an approximate match; if you set it to FALSE, VLOOKUP will return an exact match.
For example, if the service name you want to find the price for is in cell A5 and you want an exact match, the formula would look like this:
Replace “Sheet1!A1:B5” with the actual range of cells that contains your data and adjust the other values as needed.
What is also interesting about the ChatGPT model is that you can continue the conversation by adding more requests in follow up to your first one so that the application continues to troubleshoot your request until it gets clearer in providing the results you need. Now this is not necessarily foolproof as users have shown examples where the language model gets things wrong but the conversational and continuous nature of ChatGPT is amazing and to have it within our reach today is just mind blowing.
The possibilities are vast with this application. You have seen it work with emails and sheets. You’ll be surprised to know it can even do codes! Though you have to understand there are going to be bugs along the way because this is not 100% foolproof.
In summary for their usages, if you’re going through information and you want a string of results interconnected to each other much like a continuous dialogue, then ChatGPT can be helpful. For more direct information and prompts, GPT playground is faster and effective.
Through Language Model and Beyond
Though we look at ChatGPT and GPT playground predominantly from a language perspective, there’s a lot of ways in which this thing can replace what you use Google for. For example you are using Adobe Premiere Pro for video editing but you don’t want to be tied to its subscription service. So you can simply ask ChatGPT:
“what are the best free alternatives for Adobe Premiere pro”
And ChatGPT provides:
Here are some of the best free alternatives to Adobe Premiere Pro:
- DaVinci Resolve
- HitFilm Express
Each of these options have their own strengths and weaknesses, so it may be best to try a few out to see which one fits your needs best.
I mean sure, you can definitely get results from Google too, but don’t you ever have that struggle that Google search has all these different results and pages which you need to open one by one and then from there you collate the information from each of those pages to deduce the best answer. ChatGPT keeps it simple. You ask a question, it gives an answer without all the long form articles explaining minute details which at the moment you do not need. It basically filters out the information overload which search engines often produce.
Knowing what you know about how ChatGPT produces results, Google search engine will also evolve accordingly to index content that better produces results that are more user-friendly and consumable. Your content production must be more intentional rather than routinary so that Google sees you helpful in providing data that users need, minus all the flowery or technical explanations which make up long-form articles.
Important To Remember
ChatGPT and GPT playground are both very helpful, however please do not simply replace all content creation functions with these applications because their results are meant to be treated as drafts. You will notice there are now conversations about the mediocrity of blog posts created by the AI model.
We tried asking these tools to do creative things like writing advertising copy and it wasn’t really that good. However it is great for starting the process of creation, sort of like having a starting point to work with when you feel like you have a writer’s block.
ProTip: The more specific you can make a request then the better your response will be.
Here at SuperWebPros we are building a WordPress Plugin which will bring the power of GPT3 into WordPress and pre-engineer prompts for you so that way you can go ahead and just fill in key details with the already existing prompt and create desired results.
Ultimately, use the two as helping hands in optimization of your processes without necessarily building an unhealthy dependence on the services.
Do subscribe to our newsletter and we will keep you posted with exciting new developments that we may also discover along the way.