Good morning!

-TCoL

Missed our last feature article? How to Take a Day Off When You Own the Business

For its first CTV campaign, Jennifer Aniston’s DTC haircare brand LolaVie had a few non-negotiables. The campaign had to be simple. It had to demonstrate measurable impact. And it had to be full-funnel.

LolaVie used Roku Ads Manager to test and optimize creatives — reaching millions of potential customers at all stages of their purchase journeys. Roku Ads Manager helped the brand convey LolaVie’s playful voice while helping drive omnichannel sales across both ecommerce and retail touchpoints.

The campaign included an Action Ad overlay that let viewers shop directly from their TVs by clicking OK on their Roku remote. This guided them to the website to buy LolaVie products.

Discover how Roku Ads Manager helped LolaVie drive big sales and customer growth with self-serve TV ads.

The DTC beauty category is crowded. To break through, Jennifer Aniston’s brand LolaVie, worked with Roku Ads Manager to easily set up, test, and optimize CTV ad creatives. The campaign helped drive a big lift in sales and customer growth, helping LolaVie break through in the crowded beauty category.

Learn more.

By now, most small business owners have a favorite AI tool. Some use ChatGPT. Others have settled on Claude, Gemini, or Copilot. The debates about which large language model is best for which tasks are real, but they are also beside the point for most owners. The tool you already use and trust is the right starting point.

This article is not about which AI to choose. It is about how to get meaningfully more out of whichever one you already have.

Most owners who try AI and walk away unimpressed ran into the same problem: vague instructions produce vague results. AI is not a search engine. It does not reward curiosity or casual questions. It rewards specificity, context, and clear direction. The gap between a mediocre AI experience and a genuinely useful one is almost always in the quality of the instructions, not the tool.

The owners getting consistent value from AI are not using better software. They are giving better instructions, and they have learned a handful of habits that transform what the tool can do for them. Here is what that looks like in practice.

This is the single most underused technique in AI, and it costs nothing but thirty seconds. Most users open a session and jump straight to their question. The result is that AI responds as a general-purpose assistant with no knowledge of who you are, what you do, or what kind of answer would actually be useful to you.

Before you ask your first question in any session, give AI your context. Tell it who you are, what your business does, who your customers are, and how you want it to behave. Then ask it to carry that context through everything that follows.

You are an experienced small business advisor with a strong background in operations, finance, and marketing. I own a [type of business] with [number] employees. We serve [describe your customers]. When I ask you questions, factor in that context, push back when my thinking has a gap, keep your answers practical and direct, and flag anything I should verify with a professional before acting on it.

That prompt takes thirty seconds to write once and can be saved and reused at the start of every session. Everything you ask after that will be answered by an AI that understands your world rather than guessing at it. The improvement in response quality is immediate and significant.

You can also add specifics that matter to your business. If you have a particular concern, a current project, or a constraint the AI should know about, include it. The more relevant context you provide upfront, the less you will have to repeat yourself and the more calibrated the answers will be.

AI is genuinely good at producing a first draft of something you would otherwise spend 30 to 45 minutes staring at: a client follow-up email, a job posting, an auto-response for your inquiry form, a vendor introduction, a scope-of-work summary, a social media post, or a simple policy document for your team.

The key is giving it enough context. A weak prompt produces a generic result that you will spend more time editing than writing. A strong prompt produces something you can actually use with light revision.

Weak: Write me a follow-up email to a client.

Strong: Write a follow-up email to a client who requested a landscaping quote three days ago and has not responded. I run a residential landscaping company in the Southeast. The tone should be friendly but professional. Keep it under 100 words and end with a specific question that makes it easy for them to reply.

The difference in output quality is not subtle. Give AI your role, your reader, your tone, your goal, and any constraints on length or format. Treat it like a capable assistant who needs a proper brief, because that is exactly what it is. The more you invest in the prompt, the less you invest in the edit.

If you regularly receive long contracts, reports, vendor proposals, insurance policies, or email threads that require your attention, AI can summarize them quickly and pull out exactly what you need. Paste the text directly into the conversation, then ask for what matters to you.

This works well for vendor agreements before attorney review, lengthy RFPs, insurance policy summaries, email chains you need to catch up on, and meeting notes you need converted into a clean list of action items with owners and deadlines.

Example prompt: Here is a vendor agreement I received today. Summarize the key terms including the payment schedule, termination rights, auto-renewal clauses, and any provisions that limit what I can do with other vendors. Use plain language and flag anything that looks unusual or one-sided.

That last instruction, asking AI to flag unusual or one-sided terms, is worth adding any time you paste a contract. It will not replace your attorney, but it will help you walk into that conversation with better questions.

One of the highest-value uses of AI for an SMB owner is preparing for conversations with bankers, attorneys, accountants, insurance brokers, and other professionals. Most owners walk into those meetings knowing less than the person across the table. AI can close that gap in fifteen minutes.

Example prompt: I am meeting with a commercial lender next week to discuss a business line of credit for the first time. I own a five-year-old LLC in the home services industry with three employees and approximately $600,000 in annual revenue. What questions should I ask, what terms should I understand before I go, and what documents should I bring?

You can use the same approach before meeting with a new attorney, reviewing a lease, negotiating with a supplier, or evaluating any professional service you have not dealt with before. AI will give you a solid orientation, the vocabulary, and the right questions. You still bring the judgment.

AI can give you a useful orientation on topics you know little about: how a specific type of business insurance generally works, what to look for in a commercial lease, what questions to ask when evaluating a payroll provider, how a particular tax election generally operates. It is a productive first step before a professional conversation or a deeper read.

The important caveat, which belongs in every AI conversation involving facts you plan to act on, is addressed in the next section.

AI can be confidently wrong. This is not a flaw that a better tool solves. It is a structural characteristic of how large language models work, and every serious user of AI accounts for it. The systems generate plausible-sounding responses based on patterns in their training data. When the data is solid and the question is well-formed, the results are excellent. When the question touches on something specialized, recent, jurisdiction-specific, or simply outside the model’s reliable knowledge, the result can be wrong in ways that are not obvious.

The single most useful habit you can build is to follow every substantive AI response with a verification prompt. After AI gives you information you plan to act on, send this:

Now carefully review everything you just told me. Flag any statements you are not fully confident are accurate, any claims that depend on information that may have changed recently, anything that may vary by state or jurisdiction, and anything I should verify with a qualified professional before relying on it.

Or, the short version if you are time constrained:

Now carefully confirm the accuracy of all content that you just told me and provide a list of all problems and your suggested solutions.

That prompt does not make AI infallible. What it does is prompt the system to surface its own uncertainty, which it will often do honestly and specifically when asked directly. You will sometimes get a response that walks back or qualifies several things it just told you with apparent confidence. That is the system working correctly, not failing. Treat the flagged items as your follow-up list.

For anything with real financial, legal, or operational consequences, take that list to the appropriate professional. AI is an excellent starting point and a useful thinking partner. It is not a substitute for a qualified attorney, CPA, or advisor when the stakes are meaningful.

AI should not be your final word on legal, tax, or financial matters. It should not be used to generate content you will publish or send without reading it carefully yourself. It does not reliably remember context from prior sessions unless you re-establish it. And it should not substitute for your own judgment on decisions that carry real consequences for your business or your customers.

The owners who get burned by AI are almost always the ones who treat the output as finished work. The owners who benefit consistently treat it as a capable first draft or a useful starting point that still requires their eyes, their knowledge of their own business, and where appropriate, a qualified professional.

Before sending any prompt, ask yourself whether a new employee with no context about your business could follow your instructions and produce something useful. If the answer is no, add more context. That single habit, applied consistently, will improve your results more than any other change you make, regardless of which AI tool you use.

Set the stage at the start of each session. Be specific in every prompt. Verify anything you plan to act on. Those three habits will take you further than any tool comparison ever will.

This article is for informational purposes only. AI tools vary in capability and accuracy. Do not rely on AI-generated content as a substitute for professional legal, tax, or financial advice.

Have an interesting business question and need a free bit of advice? Send your question to [email protected]. No confidential info, please!