What is the best way to sustain launch over time and prevent the spike in traffic in the short run without long-term results?

First of all, don't try to prevent the short term spike! that's why you launch, to drive attention and focus for one thing in the short term. 

But how to get long term results: have a plan to continue building momentum. Like for a new product, you want to continue building momentum with content, marketing campaigns, sales team selling it, etc. And don't forget feature launches, if it's a new product, plan the announcement of the product, but then plan follow up launches for new features as you build them. 

There was a similar question below, so I encourage you to check that out, but otherwise here’s a list of other thoughts that come to mind:

• For initial launch campaigns I always try to consider what can be repurposed and reskinned post-launch for use in lifecycle campaigns, ad campaigns, and user education. That means build things in a modular way so you can take out any key dates or "brand spanking new" launch messaging without having to fully redesign collateral.
• Post launch try conducting a survey or a handful of interviews with users who have organically discovered and used the new feature/product as well as a handful of users who you thought would have discovered it organically (because they’re active users in your target audience). Figure out how people discovered and “missed it.” Plan A/B tests with more informed hypotheses.
• Set ongoing goals around feature/product awareness and adoption. If these goals aren't visible and prioritized, it'll be harder to get the resources you need to continue to invest.
• What changes can you make in your product to help increase awareness and adoption? Do you have a product tour? Do you use badges to indicate when something is new or “popular”?
• Also make sure you set expectations with leadership and stakeholders. Some releases will come and fade quickly because they aren’t big enough, they’re table stakes, or they aren’t differentiated from competitors. Ideally as a PMM, you're influencing the roadmap enough to ensure that those things are happening, but if it’s a “minimal launch,” make sure to set expectations that this product/feature not likely to hold sustained attention.
• Consider building an ongoing buzz campaign that lets you build continued credibility with this product, such as a speaker series or a sustained influencer campaign. 
• Especially if you’re making continued improvements to the product, plan a few post-launch comms to users who didn’t convert.
• Keep refreshing your creative to speak to improved benefits in your paid ads and more evergreen campaigns. No rest for the weary! It’s easy to push this to the bottom of the list, but if it’s an important product/feature, you need to keep honing and refining messaging until you see that “click” with the market.
• Work with your data analysts to gather proof points and marketing claims to your sales pages.
• Gather customer quotes and testimonials, especially if you can ask them how you compare to a competitor. In your follow-up surveys, consider asking sentiment-based questions (like CSAT) and outcome-oriented questions in a format that you can use as a proof point. For example, ask how they feel about their grades since using the product. Then you can share, "XX% of students report better grades!" 

“No, don’t do it! Don’t create a massive spike in traffic and run-up in sales that blows through your numbers.” said no one, ever.

But really, I get it, you’re not looking for empty calories, you're building a business. Here are a few thoughts.

1. Understand what might drive a short-term spike that doesn’t carry through.

Is your media spend too front-loaded?

Do you not have the right product-led growth motions to create sustainable growth?

Are you taking pricing action along with a product launch?

If you are worried about this, list out assumptions of what might cause the spike, assess the potential size and how likely it is to happen. Look at your large scale+high likelihood assumptions and evaluate the impact and how you might change your go-to-market to address it?

2. Don’t prevent a spike if you can get it, just plan for it.

Is your conversion funnel fully developed? Are you engaging non-converting traffic with customer journeys that convert deeper in the funnel? Leverage traffic spikes with additional conversion downstream.

Oh, and is the tech ready? Don’t let the site go down on the big day.

3. Don’t pull all the levers at the same time.

I know, we all get focused on the launch date and it’s exciting. In your go-to-market, play the timeline out beyond the launch moment…you can use a “rolling thunder” strategy for the message or if you have a massive email list, phase the size of audience you’re targeting.

You can also use this approach to create more focus and do fewer things bigger, over time.

4. If you get an unexpected spike, something really worked…take those learnings forward to the next campaign.

Customer stories and case studies. Proof! 

The public facing spike in the intial phase (hype creation), needs to move on to proof (value creation). 

We saw this at Helpshift where the Chatbot product started to fully automate upto 70% of a few customers' inbound customer service chats. This was way beyond what any other product in the industry was able to accomplish. 

Once the company had this data, it became very easy and effective to credibly continue to market and sell the product. It was no longer hype, the product performed, it was undeniable.